Macroeconomic Adjustment, Trade Growth: Policy Macroeconomic Model Nigeria

MARCH 1995
RESEARCH PAPER THIRTY-TWO
MACROECONOMIC
ADJUSTMENT, TRADE AND
GROWTH: POLICY ANALYSIS
USING A MACROECONOMIC
MODEL OF NIGERIA
CHARLES C. SOLUDO
ARCHIV
MIC RESEARCH CONSORTIUM
102424
JUR LA RECHERCHE ECONOMIQUE EN AFRIQUE
J
IDRC Ub
Macroeconomic adjustment, trade

and growth: Policy analysis using a
macroeconomic model of
Nigeria
This report is presented as received by IDRC from project recipient(s). It has not
been subjected to peer review or other review processes.
This work is used with the permission of African Economic Research Consortium.
© 1994, African Economic Research Consortium.

Other publications in the AERC Research Paper Series:
Structural Adjustment Programmes and the Coffee Sector in Uganda by Germina

Ssemogerere, Research Paper 1.
Real Interest Rates and the Mobilization of Private Savings in Africa by F.M. Mwega,
S.M. Ngola and N. Mwangi, Research Paper 2.
Mobilizing Domestic Resources for Capital Formation in Ghana: the Role of Informal
Financial Markets by Ernest Aryeetey and Fritz Gockel, Research Paper 3.
The Informal Financial Sector and Macroeconomic Adjustment in Malawi by C. Chipeta

and M.L.C. Mkandawire, Research Paper 4.
The Effects of Non-Bank Financial Intermediaries on Demand for Money in Kenya by

S.M. Ndele, Research Paper 5.
Exchange Rate Policy and Macroeconomic Performance in Ghana by C.D. Jebuni, N.K.
Sowa and K.S. Tutu, Research Paper 6.
A Macroeconomic-Demographic Modelfor Ethiopia by Asmerom Kidane, Research Paper

7.
Macroeconomic Approach to External Debt: the Case of Nigeria by S. Ibi Ajayi,

Research Paper 8.
The Real Exchange Rate and Ghana's Agricultural Exports by K. Yerfi Fosu, Research
Paper 9.
The Relationship Between the Formal and Informal Sectors of the Financial Market in
Ghana by E. Aryeetey, Research Paper 10.
Financial System Regulation, Deregulation and Savings Mobilization in Nigeria by A.

Soyibo and F. Adekanye, Research Paper ii.
The Savings-Investment Process in Nigeria: an Empirical Study of the Supply Side by A.
Soyibo, Research Paper 12.
Growth and Foreign Debt: the Ethiopian Experience, 1964-86 by B. Degefe, Research
Paper 13.
Links Between the informal and Formal/Semi-Formal Financial Sectors in Malawi by C.
Chipeta and M.L.C. Mkandawire, Research Paper 14.
The Determinantsof Fiscal Deficit and Fiscal Adjustment in COte d'Ivoire by 0. Kouassy and B.
Bohoun, Research Paper 15.
Small and Medium-Scale Enterprise Development in Nigeria by D.E. Ekpenyong and M.0.
Nyong. Research Paper 16.
The Nigerian Banking System in the Context of Policies of Financial Regulation and Deregulation
by A. Soyibo and F. Adekanye, Research Paper 17.
Scope, Structure and Policy Implications of Informal Financial Markets in Tanzania by M. Hyuha,
.0. Ndanshau and J.P. Kipokola, Research Paper 18.
European Economic Integration and the Franc Zone: the Future of the CFA Franc after 1996. Part
I: Historical Background and a New Evaluation of Monetary Co-operation in the CFA
Countries by Allechi M'bet and Madeleine Niamkey, Research paper 19.
Revenue Productivity Implications of Tax Reform in Tanzania by Nehemiah E. Osoro, Research

Paper 20.
The Informal and Semi-Formal Sectors in Ethiopia: a Study of the Iqqub, Iddir and Savings and
Credit Co-operatives by Dejene Aredo, Research Paper 21.
Inflationary Trends and Control in Ghana by Nii K. Sowa and John K. Kwakye, Research Paper
22.
Macroeconomic Constraints and Medium-Term Growth in Kenya: A Three-Gap Analysis by F.M.

Mwega, N. Nguguna and K. Olewe-Ochilo, Research Paper 23.
The Foreign Exchange Market and the Dutch Auction System in Ghana by Cletus K. Dordunoo,
Research Paper 24.
Exchange Rate Depreciation and the Structure of Sectoral Prices in Nigeria Under an Alternative
Pricing Regime, 1986-89 by Olu Ajakaiye and Ode Ogowu, Research Paper
25.
Exchange Rate Depreciation, Budget Deficit and Inflation - The Nigerian Experience by F.
Egwaikhide, L. Chete and G. Falokun, Research Paper 26.
Trade, Payments Liberalization and Economic Performance in Ghana by C.D. Jebuni, A.D. Oduro
and K.A. Tutu, Research Paper 27.
Constraints to the Development of Non-Traditional Exports in Uganda, 1 981-90

by G. Ssemogerere and L.A. Kasekende, Research Paper 28.
Indices of Effective Exchange Rates: A Comparative Study of Ethiopia, Kenya and the Sudan by
Asmerom Kidane, Research Paper 29.
Monetary Harmonization in Southern Africa by C. Chipeta and M.L.C. Mukandwire, Research

Paper 30.
Tanzania's Trade with PTA Countries: A Special Emphasis on Non-Traditional Products by Flora
Mndeme Musonda, Research Paper 31
Macroeconomic adjustment, trade
and growth: Policy analysis using
a macroeconomic model of
Nigeria
Charles Chukwuma Soludo

University of Nigeria, Nsukka
Nigeria
AERC Research Paper 32

African Economic Research Consortium, Nairobi
March 1995
© African Economic Research Consortium, 1995
Typeset by the African Economic Research Consortium,

P.O. Box 62882, Nairobi, Kenya.
Printed by The Regal Press Kenya Ltd.,

P.O. Box 46166,
Nairobi, Kenya.
ISBN 9966-900-26-8
Contents
List of tables
List of figures
Acknowledgements
Abstract
I Introduction
II Theoretical model and estimation results
III Policy simulations and analysis of results 48
IV Conclusions and policy issues 55
Notes 62
References 64
Appendix 1: Data Sources and Limitations 69

Appendix 2: List of Variables
Appendix 3: Estimation Results 74
Appendix 4: Figures Plotting Simulation Results 82
List of tables
1. Some economic and financial indicators 6
2. Exports by major commodity groups 8
3. Exports by major trading partners 8
4. Imports by major groups 9
List of figures
1. Flowchart 18
Acknowledgements
I am indebted to a number of people who have provided extremely helpful comments and
assistance during the course of this study. In particular, I would like to thank Benno
Ndulu (AERC) and Warwick McKibbin (Brookings Institution) for their valuable
comments on the earlier specifications of the model. I am especially grateful to Ralph
C. Bryant for his extensive comments and suggestions on early drafts of this report.
Some aspects of the research were begun while I was Visiting Fellow at the Brookings
Institution (1991-92); I would like to thank the institution for the necessary assistance.
More so, I thank the Brookings Institution for providing office space and facilities, and
Ralph Bryant and Phil Bagnoli for their technical assistance during the trial simulations
of the model in April/May 1993. I would also like to thank Julius Asogu (Central Bank
of Nigeria) and Adeola Adenikinju (University of Ibadan) for their assistance at various
stages of the data collection, and Abiy Zegeye for excellent research assistance. Finally,
I would like to thank the participants at the AERC research workshops for their
comments.
Abstract
This research paper presents a medium-sized macroeconomic model of the Nigerian
economy that incorporates both forward-looking, model-consistent expectations and,
backward-looking, adaptive expectations in the assets and goods markets. The model is
designed for use in evaluating alternative policy regimes under the structural adjustment
programmes in Nigeria. It is shown to have sensible short- to medium-term simulation
properties, and can be applied to analysis of a wide range of policy issues of the 'what
if kind. In this report, only demonstrative simulations are run to shed light on the
implications of alternative expectations regimes, and domestic monetary and fiscal
policies under SAP for Nigeria's trade performance and growth prospects in the 1990s.
It is shown that key proposals for 'fiscal discipline' and enforcing a regime of
'stabilization with sterilization' in respect of windfalls from oil exports are consistent with
the other SAP objectives of promoting macroeconomic stability, trade and growth in the
medium and long runs.
I Introduction
Introduction/statement of the problem

Macroeconomic stability and growth are the two principal objectives of the World
Bank/IMF supported Structural Adjustment Programmes (SAP). These programmes
constitute a logical response to the debt and development crises experienced by most of
the developing countries, especially since the early 1980s. The specific contents of the
programmes, the choice of instruments and the timing and sequencing of the
implementation depend or should depend on the stage of development of a country's
economic institutions, the size of its debt burden, the nature of the existing
macroeconomic imbalances and policy distortions and political sensitivities. In a broad
sense, the programmes often involve the integration of traditional short-term stabilization
essentially the correction of external and internal imbalances through aggregate demand
management with medium to longer-term structural measures aimed at stimulating the
supply side of the economy.
The basic elements of the programmes generally involve:(i) Monetary restraint aimed
at reducing the growth of absorption and the rate of inflation; (ii) Interest rate policies
aimed at keeping real interest rates positive but low; (iii) Fiscal restraint to reduce the
fiscal deficit to a sustainable level and thereby restrain aggregate demand pressures; (iv)
Exchange rate action to ensure a real exchange rate that improves international
competitiveness and creates the incentive for expanding the production of internationally
traded goods; (v) External financing policies to reduce the stock of external debt if it is
perceived to be currently unsustainable, or to limit foreign borrowing if it is likely to
become so in the future; (vi) Structural reforms (such as financial sector reforms,
producer pricing policies, trade liberalization, and tax reforms) to make the economy
flexible and efficient1.
About 30 countries in Africa began implementing SAPs in the early 1980s. Their
experiences have varied, depending on the peculiar structure of the economy, and the
nature of their macroeconomic imbalances. In Nigeria, for instance, the results of SAP
policies since their introduction in 1986 have not been salutary. In spite of the conscious
efforts to implement most aspects of the basic elements of SAP as outlined above,
Nigeria's external debt has increased from about $18 billion in 1986 to over $33 billion
in 1992. Government's annual budget-deficit has exceeded 10% of GDP, price inflation
has been very high, and capacity utilization has been very low even though the annual
growth rate of GDP has been modest. Controversies have raged among policyanalysts
2 RESEARCH PAPER 32
and policymakers as to what went wrong, and what other viable macroeconomic policies
could be implemented to ensure the success of structural adjustment. In the course of the
debates, and in the light of the structure of the economy, the nature of external shocks
that affect it and the results of the adjustment process so far, certain issues and questions
have been raised about the prospects of SAP policies in Nigeria:
1. Given the nature of such external shocks, the structure of the Nigerian economy and
its fiscal federalism, can the government realistically avoid persistent budget deficits
without sacrificing the growth objectives of the adjustment programme?
2. In view of the institutional practice of active monetization of foreign exchange

earnings, persistent budget deficits in the face of rising debt interest payments and
moderate current account surpluses, can the Central Bank of Nigeria successfully
control money supply and inflation, and hence stabilize the exchange rate? Are there
significant crowding-out effects of the public sector deficit financing on the private
sector?
3. To what extent can the liberalization of the foreign exchange market, and thus the
expenditure-switching effects stemming from exchange rate changes, reduce the stress
on the current account and also boost the non-oil export sector? Furthermore, given
the structure of Nigeria's tradeable goods sector, what are the prospects of the external
sector (mainly exports) leading the economy to sustainable growth in the medium to
long-term?
4. In the absence of debt forgiveness or equivalent operations, and in the face of

declining oil prices, are the present adjustment processes sustainable in terms of
ensuring macroeconomic stability with growth even in the long run? Or are the basic
elements of stabilization and liberalization in conflict, therefore requiring, as Sachs
(1987) argues, 'different sequencing and timing procedures between stabilization and
liberalization measures'?
5. What other structural and institutional reforms should the government embark on
to aid a 'growth-oriented-adjustment' process?
The foregoing are certainly issues and questions that are central to a critical evaluation
of the on-going adjustment efforts in Nigeria. Such an evaluation exercise, or even the
design of the adjustment programme for that matter, needs a formal model framework,
either implicitly or explicitly. An explicit, formal model framework, in spite of its
shortcomings, is preferred to implicit models. A quantitative model, for instance, not
only provides a theoretical structure for understanding the linkages among principal
macroeconomic variables but also provides, in broad order of magnitudes, a systematic
MACROECONOMIC ADJUSTMENT, TRADE AND GROWTH 3
sense of the signs and magnitudes of the relevant parameters and the transmission of
shocks from one sector to others and the feed-back effects to the originating sector. In
essence, the dynamic properties of the adjustment process can be meaningfully studied
with formal models.
This study is therefore motivated by two driving needs. The first is the absence of an
operational macroeconomic model of the Nigerian economy that captures the current SAP
environment. A major objective of this research is to develop such a model. The second
need is to use the model to evaluate the internal consistency of some of the proposals for
fiscal and monetary discipline in Nigeria. In particular, we use some standardized shocks
to evaluate the potentials of budget-deficit reduction policy in promoting macroeconomic
stability, trade and growth in the short and medium runs. Also, the implications of a
positive oil price shock under a regime of stabilization with sterilization are evaluated.
This is designed to shed light on the consistency of the proposals for oil stabilization
account and sterilization of oil windfalls with the objectives of SAP in Nigeria.
Objectives of the research agenda

The basic, short-run goals of the research agenda are to build and maintain a general-
equilibrium economic model of the Nigerian economy that captures the essential elements
of the current SAP environment in Nigeria. Over the medium and longer-runs, the
model is to be maintained alongside, and used interactively with, some global models
(such as the IMF's MULTIMOD) so as to permit the simulation of the dynamic paths of
the external shocks that affect the Nigerian economy. The Nigerian model, together with
its ad hoc linkages to the global models, is to be maintained for policy analysis of the
Nigerian economy, and could be applied to illuminate a variety of important policy
issues. Over a longer-run, global empirical models such as MULTIMOD will be further
refined. It will ultimately be desirable to disaggregate MULTIMOD's current developing
country region (DC) into several separate developing country regions, including an
African region. The more progress that is made in successfully disaggregating the
MULTIMOD DC region, the more it will be possible to refine estimates of the linkages
between the Nigerian economy and the rest of the world. Alternatively, the Nigerian
model may provide an important starting point toward a future construction of a sub-
regional model of the Economic Community of West African States (ECOWAS) or even
an African regional model.
The specific objectives of the present research are:
1. to build a medium-sized macroeconomic model of the Nigerian economy that

captures the basic elements of the SAP policy environment to verify the model's
consistency and statistical properties, and to simulate it under alternative assumptions
4 RESEARCH PAPER 32
about the expectations formation process of private economic agents in Nigeria

(adaptive and model-consistent expectations).
2. to use standardized shocks that are consistent with various fiscal and monetary
policy proposals under SAP to shed light on the implications of these proposals for
Nigeria's trade, economic growth and macroeconomic stability in the short and
medium runs.
3. to make policy recommendations on the basis of the simulation results.
Overview of major macroeconomic trends during the

SAP regime in Nigeria
For several years prior to the introduction of SAP policies in 1986, the Nigerian
economy was in deep crisis. This was precipitated mainly by the collapse of the oil
prices and the gross mismanagement of the economy in general. The oil boom era of the
of mid 1970s to 1980 witnessed a massive expansion in the size of government and most
of the oil revenues were spent to expand physical infrastructure; finance increases in
recurrent expenditures resulting mainly from the sharp rise in wages and salaries,
expansion of the import-based consumer goods industry and import-substitution
industrialization programmes; and inefficient and sometimes bogus manufacturing
industries. The policy direction also led to the deterioration of the agricultural sector and
non-oil export sector, and maintained a highly overvalued exchange rate. Price controls
were generally enforced and both quantitative and tariff-based import restrictions were
frequently applied to correct perceived imbalances in the current account.
Following the collapse of the oil prices, export receipts declined from $26 billion in
1980 to $9.4 billion in 1989. In real terms, oil export revenues fell below their pre-oil
boom era, and the resulting external and fiscal imbalances became unsustainable. The
current account deficit reached 6% of GDP in 1983, and the fiscal deficit was double that
figure. Between 1981 and 1986, real aggregate demand fell by 35%. The situation
continued to worsen and in 1989, per capita GDP had dropped to around $250 compared
to $1090 in 1980. During the oil boom era, the major mechanism of government's
patronage was public expenditure; during the slump this source was replaced by rents
from foreign exchange allocation. An indication of the growth in such rents and from
the implicit tariff rates generated by import restrictions can be obtained by examining the
parallel market premium over the official rate. In early 1981, this premium was only
37%; during 1983 it exceeded 200% and by 1986 it was 330% (Bevan, et al, 1992:23).
No doubt this policy reduced the income of the export sector - a cumbersome and
MACROECONOMIC ADJUSTMENT, TRADE AND GROWTH
rationed access to imports was imposed - and thus handicapping firms that depended on
imported inputs. It was obvious to policy analysts that urgent policy responses were
needed but up to the mid-i 980s, there was no consensus as to what should be done.
When the Babangida regime came into office in August 1985, it was faced with even
deeper economic crisis. Oil prices crashed to their all-time low level, and by the time
the government had spent one year in office, real income had fallen by more than 20%,
an unprecedented decline even by the harsh standards of the preceding years. The
inevitability of a fundamental economic restructuring policy response became urgent and
the Babangida regime embarked on the World Bank/IMF supported structural adjustment
programme in September 1986. The most important elements of the adjustment
programme were:
• abolition of the import-licensing regime and the liberalization of the foreign

exchange market;
• rationalization and restructurLg of tariffs in order to aid industrial diversification;
• introduction of measures to stimulate domestic production and broaden the supply

base of the economy;
• deregulation of the monetary and financial system while aiming to maintain tight
control over budgetary deficits;
• substantial liberalization of trade and payments;
• adoption of appropriate pricing policies, especially for petroleum products and

public enterprises; and,
• encouragement of rationalization and privatization of public sector enterprises.
It is one thing to articulate sound economic policies; it is another to effectively

implement them. Effectiveness in the implementation of specific adjustment programmes
by the government is crucial in determining general economic performance. This is
especially true in the peculiar circumstances of such economies as Nigeria, where the
public sector is the dominant economic agent. The mainstay of the economy - the oil
sector - is owned by the federal government and accounts for over 90% of Nigeria's
foreign exchange earnings and over 70% of government revenues, and represents about
13% of the GDP. Almost all of Nigeria's external debts are the obligations of the federal
government, with a large proportion owed to external private creditors at variable interest
rates. Nigeria is a price taker in both the oil and the primary commodity (agricultural
export) markets. In the case of oil, the quantity of output is also exogenously determined
6 RESEARCH PAPER 32
by OPEC. These characteristics of the public sector have important implications for the
success of the adjustment process.
Table 1: Some economic and financial indicators
GOP % Ex. Rate Inflation Lending Mini. M1% GOV. CBN/ DDBT
Naira/ rate redisc ount TOT GOV %
US$ (prime) rate
1986 3.13 1.73 5.40 17.50 10.00 5.23 0.53 0.85 1.79
1987 1.70 3.97 10.20 20.50 13.86 13.70 0.53 0.73 29.31
1988 4.20 4.54 38.30 18.40 12.75 41.90 0.52 0.84 27.83
1989 5.20 7.37 50.50 25.70 18.50 21.51 0.41 0.92 21.31
1990 5.20 8.35 7.50 26.50 18.50 44.90 0.45 0.80 47.40
1991 4.60 10.87 13.00 21.00 15.50 32.60 0.46 0.89 38.20
1992 4.10 19.47 44.60 31.20 17.50 66.40 0.52 0.87 39.00
Sources: Computed from Central Bank of Nigeria's Annual Report and Statement of Accounts
(various years)
CBN's Statistical Bulletin, 1992
Notes:
GDP = Growth rate of GDP

CBN Total credit granted by the Central Bank of Nigeria
DDEBT = Domestic debt of the Federal government (growth rate)
TOT = Total credit to the domestic economy by the banking system
A review of some economic indicators shows that the adjustment programme has had
mixed results in Nigeria. As Table 1 shows, the annual growth rate of the GDP (at 1984
constant factor cost) has been modest, recovering from a mere 1.7% in 1987 to an
average annual growth rate of over 4% since 1988. This appears an impressive
performance especially when the general macroeconomic instability occasioned by
government's fiscal and monetary policies is considered. For example, as is evident from
Table 1, the government has not been successful in controlling the growth rate of money
stock and consequently the inflation rate has been quite high over the years. The narrow
money stock has been growing quite rapidly, reaching an unprecedented high rate of 66%
in 1992. These increases in monetary aggregates are often blamed on the unsustainably
high rate of monetization of foreign exchange earnings of the public sector and the sharp
rise in credit to the domestic economy. High growth in credit to the domestic economy
was attributed to the huge federal government deficit, which was mainly financed by the
Central Bank of Nigeria. The federal government's budget deficit has been around 10%
of GDP for most years of the adjustment programme. As also indicated in Table 1, the
banking system's credit to the government as a proportion of total credit to the domestic
economy has exceeded 40%, while the proportion of government's total credit financed
by the Central Bank has increased from about 55% in 1985 to about 90% in 1991.
Correspondingly, government's total domestic debt obligations have been growing at very
high rates.
It should be observed that in most of the years since the adjustment programmes
started in 1986, the annual budgets proposed by the federal government have often been
in slight surplus or, at worst, in balance, perhaps in order to keep to the requirement of
'fiscal discipline'. But also, in most years, the ex post annual budgets have tended to be
in huge deficits. This is often attributed to, among other things, the unanticipated
increases in interest payments on the government's external debt and! or shortfalls in oil
revenues as a result of unanticipated falls in oil prices. According to the Central Bank
of Nigeria's Annual Report and Statement of Accounts (1992, p.3):
The larger deficit, in terms, was due mainly to large external

debt service payments, as well as on the execution of the political
programmes and economic development projects. Public debt
outstanding increased substantially by 58.9 percent over the 1991
level.... Of the total amount expended, debt service payments
constituted 61.1 percent, while expenditures on non-debt recurrent and
capital expenditures absorbed 38.9 percent. Consequently, government
fiscal operations resulted in an overall deficit of N44,158.5 million or
9.8 percent of GDP which was financed mainly by the Central Bank
of Nigeria.
The huge budget deficits and the consequent rapid monetary expansion exerted
pressures on the naira exchange rate, which depreciated persistently. Indeed, between
January and March 1992, the naira depreciated by over 60% against maj or currencies.
Domestic market interest rates have continued to rise since the adjustment programme
due, in part, to the liberalization of the financial sector, and in spite of the excessive
growth of money stock. It might seem paradoxical that interest rates should be rising
instead of falling when budget deficits are financed by printing money, which generally
increases the liquidity of the economy. In Nigeria however, the only handle that the
monetary authority has in controlling the money supply is the credit to the private sector.
Thus, given the helplessness of the monetary authority to control government's overdraft
with the Central Bank, and in an attempt to stem the growth of the money supply, it tries
to peg the maximum growth rate of credit to the private sector. As banks try to make
profits from the limited lending activities permitted by the Central Bank, the spread
between the deposit and lending rates continues to widen. The rising interest rates
combined with the rapidly depreciating local currency and high inflation have tended to
slow the growth of output, especially in the non-oil sectors.
8 RESEARCH PAPER 32
Table 2: Exports by major commodity groups
Agricultural Mineral Manufacture! Textiles Chemicals Other Other

products products semi manufacture manufa ctureEx.
(oil) of products exports
1986 4.6 93.8 0.6 1.0

1987 5.2 92.9 0.2 1.6
1988 9.1 88.4 0.3 2.2
1989 3.2 94.9 0.2 0.2 0.0 0.1 1.4
1990 2.1 97.0 0.1 0.1 0.1 0.4 0.2
1991 2.7 96.2 0.1 0.2 0.2 0.1 0.4
Source: Compiled from Central Bank of Nigeria's Annual Report and Statement of
Accounts (various years)
One major achievement of the SAP regime was the elimination of import-licensing and
the bold steps to reform the structure of tariffs. Almost all the price controls and import
licensing requirements were abolished and the import prohibition list was shortened. The
30% import surcharge was abolished and the government also adopted an interim import
duty and excise schedule that somewhat reduced the level of protection. The reform of
the foreign exchange market and the consequent devaluation of the naira was expected
to have beneficial effects on the external sector, especially on the exports of non-oil
products. As Table 2 shows, the non-oil exports (agricultural products) as percentage
of total exports increased steadily from 4.6% in 1986 to 9.1 % in 1988, before declining
to 2.0% in 1992.
Table 3: Exports by major trading partners (percentage of total exports)
Common Eastern Japan UK USA - ECOWAS Western Others

Wealth Europe Europe
Countries
1985 5.99 0.28 0.06 4.60 18.06 3.31 61.58 6.12

1986 4.29 0.23 0.15 5.67 34.96 3.82 47.58 4.19
1987 2.89 0.03 0.08 1.79 46.99 5.53 40.07 2.63
1988 3.00 0.19 0.12 1.89 45.97 5.57 38.61 5.35
1989 0.85 0.10 0.22 1.75 53.32 5.52 34.36 3.88
Source: Central Bank of Nigeria's: Statistical Bulletin, 1992.

Oil exports continue to dominate the export products. Indeed, it is believed that the
increase in the percentage of non-oil exports from 1986 to 1988 may be due to a slump
in the oil market in those years rather than to any appreciable increase in non-oil exports.
In terms of direction of Nigeria's exports, the structure has not changed. Table 3
indicates that the United States remained Nigeria's major trading partner. Since 1987,
the USA has imported more from Nigeria than the Western European countries
combined. The US imports from Nigeria consist almost entirely of oil.
The trend that started in the import structure in the early 1980s continued throughout
the SAP period. This trend is reflected by the domination of the import items by the
capital goods and raw materials. The consumer goods share of total imports declined
over the period, but showed a sharp rise to 34% in 1992 (see Table 4). On the whole,
the current account balance has been modestly positive especially since 1989. The
negative balance of payments accounts is mainly due to the impact of debt service
payments on the capital accounts.
Table 4: Imports by major groups (jercentage of total imports)
1985 1986 1987 1988 1989 1990 1991
Consumer goods 22.1 21.8 24.1 28.7 27.3 26.7 24.8 33.8
Durable goods 7.8 3.6 4.7 1.0 4.0 3.20 4.20 3.2
Non-durable goods 14.3 18.2 19.4 27.7 23.3 23.5 20.60 30.6
Capital goods!
Raw materials 76.3 78.1 75.9 71.3 72.6 73.6 74.8 65.6
Capital goods 34.9 36.2 42.4 32.0 44.7 40.5 38.0 31.7
Raw materials 41.4 41.9 33.5 39.2 27.9 32.8 36.8 33.9
Miscellaneous 1.6 0.1 0 0.0 0.10 0.0 0.4 0.6
Source: Central Bank of Nigeria, Annual Report and Statement of Accounts (various
years).
From the brief review of the adjustment process so far, it is evident that there are
therefore two main sources of shocks originating externally whose impacts on the
Nigerian economy, and hence on the outcomes of the adjustment process, are substantial.
One is the variation in the interest rates on the government's debt to foreign creditors.
The other is the variation in oil prices, which is sometimes unanticipated. These two
shocks have strong impacts on the government's budget behaviour. Tangential to the
second shock is the terms of trade shock, especially since the early 1980s when the prices
10 RESEARCH PAPER 32
of most of Nigeria's exports began to decline in the face of progressive increases in the
prices of its imports. In an economy that is heavily import dependent, the balance of
payments implications of the declining terms of trade are severe. More so, with this
dependence of both the tradeable and non-tradeable goods sectors on imported inputs, the
potential short-to medium-term costs of an adjustment process that relies substantially on
expenditure switching and reducing policies may be great.
II. Theoretical model and estimation results
The last section showed that Nigeria's macroeconomic trends are not sustainable and
reviewed some desired changes in policy regime. The benchmark model presented here
is not designed to exactly capture past trends. Rather, the model combines some
characteristics of the economy with policy rules that are consistent with proposals in the
SAP package or that are consistent with policy rules that should be adopted. The
benchmark model is largely counterfactual to historical experience and thus the model
asks, 'what would happen if' Nigeria were to have a policy regime in place such as the
one assumed in the benchmark model.
Clarification of methodological issues

Modelling a small, open developing economy such as Nigeria presents some analytical
and methodological problems. The issues that confront the model-builders include the
choice of an appropriate theoretical framework for modelling the economy; the extent of
institutional and structural details to be incorporated in such a model; and what should
be the 'correct' use of macro models in such economies.
The state of the economics profession indicates that there are several modes of
economic reasoning, based on alternative visions of the economic system or specified
analytical techniques. The different visions of the economic system, formulated into
alternative theories of economic behaviour, constitute the different analytical frameworks
for economic analysis. Thus, even when the major objective of a modelling activity is
the systematic understanding of the structure and operations of an economic system, the
diverse theoretical frameworks about the functioning of the economy significantly
determine the specifications of the equation system and, implicitly, the purposes the
model can serve. The choice of a theoretical framework is therefore critical to a model's
construction and interpretation.
In many developing countries especially in Africa, economy-wide models as inputs into
policy analysis are still at a rudimentary stage of development. Most of the models are
developed as dissertations for doctorate degrees and are hardly maintained for policy
analysis. In a majority of the cases, the analytical transparency of model design and
results are not paramount considerations. The emphasis is often on building models that
'track history' and apply mainly to short-run policy analysis. One justification for this
strategy is that these developing countries are characterized by rudimentary financial
systems, dual economic systems, structural rigidities, and economic activities heavily
1 2 RESEARCH PAPER 32
controlled by the government. Also, economic policy in these environments is perceived

to be .ad hoc in nature. Policy makers are not guided by any consistent framework, nor
is intertemporal consistency in economic behaviour emphasized. In view of this, many
scholars have questioned the relevance of any of the mainstream analytical frameworks
in explaining economic behaviour in most developing countries. Modelling efforts have
therefore largely proceeded by specifying equations that, given the assumed peculiar
institutional characteristics of the economy, appear to give 'sensible' explanations of
short-run economic behaviour. Long-run sustainability of short-run behaviour is not
emphasized.
A large number of such macroeconomic models have been constructed for Africa. A
survey of modelling efforts in Africa by Harris (1985) showed that there were about 184
macroeconomic models on various African economies, 33 of which were on Nigeria.
Most of the models on Nigeria (see for example, Ojo, 1972; UNCTAD, 1973; Adamson,
1974; World Bank, 1974; Olofin, 1977; Uwujareen, 1977; NISER, 1983; Ekeoku, 1984;
CEAR-MAC IV [see Olofin, 1985]; Poloamina, 1986; Soludo, 1989; etc.) are either the
products of doctoral theses or represent one-shot research efforts to write journal articles,
or the models are built to analyse specific issues and not maintained thereafter. Thus,
with the exception of the CEAR model MAC-IV (developed and maintained by the staff
of the Centre for Econometric and Allied Research, University of Ibadan), none of the
models has been maintained for policy analysis2.
Despite the large number, one sobering observation by Harris (1985:4) was that very
few of the models in Africa shed any light on any of the issues that have been central to
international discussions of IMF and World Bank proposals for stabilization and structural
adjustment. He notes in particular that the consequences of exchange-rate adjustment,
relaxation of quantitative controls over imports and credit allocation, raising domestic
interest rates, controlling budget deficits, etc., are barely touched upon in any of the
models that have actually been used by decision makers. Harris concludes by noting that
one of the major challenges facing macroeconomists in Africa has to do with the
development of appropriate theory that reflects understanding of how particular
economies work and using this as a guide in the specification of models that are designed
to shed light on the effects of alternative policy actions.
This is the crux of the matter and points out an important challenge for Nigerian
macroeconomic modellers in particular. In Nigeria, for instance, macroeconomic
modelling tradition is largely predicated on the old policy regime of excessive
governmental controls and without a consistent analytical framework for either policy
analysis on modelling exercises. Equation specifications have then been an exercise in
the search for those that can give better explanatory power, rather than rooted in any
framework of economic behaviour. The emphasis of the models is on 'tracking history',
and the long-run sustainability of policy actions are largely ignored. Thus, the
intertemporal budget constraints are not observed, and these models are not robust for
analyzing the consequences of major policy shifts.
MACROECONOMIC ADJUSTMENT, TRADE AND GROWTH 1 3
Since 1986, Nigeria has been implementing the IMF/World Bank supported structural
adjustment programme. Various short-to long-term measures designed to deal with the
country's severe balance of payments disequilibrium and macroeconomic instability have
included standard stabilization policies (mainly expenditure-reducing in nature) and
growth-inducing, supply-side reforms. These supply-side reforms involve significant
liberalization of trade, capital inflows/outflows, financial markets and foreign exchange
markets. Economic policy appears to follow a discernible pattern as dictated by the
analytical framework for standard SAPs in developing countries (see for example,
Research Department, IMF 1987; Corbo, et al (eds), 1987; Bacha and Edwards, 1988;
etc). Domestic economic behaviour in response to changes in fiscal, monetary and
exchange rate policies appears to be altering. Methodologically, therefore, this effort
differs from earlier ones in the important sense that the modelling exercise is predicated
on the SAP framework. The basic assumptions that underlie the modelling strategy are
those that are consistent with the policy reforms under the SAP regime.
Our point of emphasis and departure from existing models is the need to root the
modelling exercise in economic theory. The sustainability of a model across generations
of users requires that it be understandable by the people who turn the crank and the
ultimate users of the output of the model. This requires that the model be firmly
grounded in economic theory so that the output of the model is intuitive to the user, or
if some results are not initially intuitive because of general equilibrium outcomes that are
not apparent in the partial models we all use in our heads, then it should at least be
understandable (McKibbin, 1991:1). The basis for our predilection for the analytical
transparency of the model design and interpretation is the current state of uncertainty
about the functioning of a developing economy under a SAP regime. When uncertainty
about a subject is pervasive, the profession should value analyses that can be readily and
clearly understood (Bryant, 1992).
One may argue, and there may be ample reason to do so, that the received theories do
not work well in the African context.
There is no agreement among modellers about the correct stance of theory in policy
analysis and modelling: while some place greater emphasis on economic theory as the
driving force, others emphasize the structural features of the economy. The view by
some analysts is that economic theory is economy-specific, and thus the structural and
institutional characteristics should be taken as a given, and economic theory and policy
making designed to be consistent with the given features. It is therefore expected that
a modelling exercise would aim to capture as much of the institutional and structural
details as possible even if the analytical transparency of the results is suspect. Such
models can indeed be important, depending on the purpose of the exercise in the first
instance. For instance, an essential feature and emphasis of such models is to track
history and explain reality as determined by the characteristics of the economy. In an
environment of stable policy regimes and structures, the predictive ability of such models
would be high. The robustness of the model becomes suspect if major regime shifts
occur (as under SAP) that also significantly alter the institutional and structural
characteristics of the economy. In that instance, the past trends are less useful in
understanding future prospects.
An alternative perspective assumes the universality of economic theory. It is argued
that if the theories fail to work in some circumstances, it is because of certain distortions
in the institutional and structural framework. In this case, both the institutional and
structural characteristics are creations of economic policy and thus a theory-consistent
policy framework can alter the given structures. For modelling purposes, therefore, this
perspective places paramount importance on the analytical consistency of the model
design.
In practice, however, economic modelling entertains the two assumptions. The
difference is often on emphasis, and the choice of emphasis entails some trade-off.
Unfortunately, the trade-off seems very steep: to obtain a highly disaggregated model
capturing details of the institutional and structural characteristics, it appears to be
necessary to sacrifice a great deal of analytical rigour; and to obtain a model whose
results are clearly intuitively interpretable, some of what is apparently the reality is
sacrificed.
Our preferred strategy for this research is to entertain as much of the institutional
characteristics of the Nigerian economy as necessary while still ensuring that the resulting
model is analytically intuitive. This is primarily because of the focus of the model as a
policy model that can be used to analyse the implications of alternative policy scenarios
of the what if kind: ultimately, the model can be revised to analyse other important policy
issues, such as the concerns of the policymakers about stabilization with or without
sterilization, nominal income or money targeting by the monetary authorities, exchange
rate regime issues, alternative fiscal rules with respect to taxation, borrowing, debt
servicing and spending, etc. We believe that these kinds of scenario analyses can be
more meaningfully done with a model whose results are intuitive.
The model reported here is highly aggregative in nature. The issue of the extent of
sectoral disaggregation to entertain in a model is an important methodological issue, but
one that is largely informed by the purpose of the study. Our strategy in the current
effort is to start with a relatively small model whose results can be understood and, later
on, to systematically expand the model structures.
For example, there are several aspects of the model equations where we consider
further disaggregation necessary. The production block, exports, investment expenditures,
and prices are some of the variables that may deserve further disaggregation depending
on the use to which the model would be put. Of course, there is no rigid procedure for
any disaggregation exercise nor is there any limit to the conceivable kinds of
disaggregation that can be entertained. For example, the production or supply sector
could be disaggregated into say, agriculture, livestock, mining, manufacturing, utilities,
building, construction, distribution, transport, communications, government, and other
services. Many of the above can, in turn, be sub-divided into smaller sub-sectors.
Exports can also be disaggregated by major trading partners (that is, by direction of trade
instead of commodity groups) but this is not pursued in the present model because one
of our objectives is to examine the performance of non-oil exports when macroeconomic
policies change. In a later version of the model when an effort would be made to refine
the model's linkages to some global models, there may be some spin-offs in
disaggregating exports by trading partners. However, given the explicit objectives of this
study, we doubt that the qualitative results will be significantly improved by any further
ambitious disaggregation scheme.
Two versions of the model are created: one assumes model-consistent, forward-looking
expectations, and the other assumes that economic agents are backward-looking (adaptive)
in their expectations . The model can also be simulated by switching between either a
flexible or fixed exchange rate regimes. The version reported in this study assumes a
flexible exchange rate regime. It is important to note here that the nature of transmission
channels and links across model blocks is determined primarily by the assumed fiscal and
monetary policy rules in place. This is one aspect of the modelling strategy that differs
markedly from earlier practices in most models of African economies. Here, we make
explicit assumptions about the type of fiscal and monetary policy rules, and the
sustainability of such actions is also of prime importance, since the model is crucially
concerned about the sustainability of adjustment paths. For example, one may expect that
in a model of Nigeria, the links between the monetization of budget deficits, price
inflation and real exchange rate should be evident. In essence, money supply may be
expected to be endogenous, and the exchange rate pegged to the US dollar. This is one
possible version of the model that can be created, but the links may be somewhat
different principally because of the assumed fiscal rule. The discussion of alternative
fiscal closure rules (see specification of the government sector) aptly summarizes the
basic issues involved. During simulations, one can impose some of these fiscal rules and
examine their implications on the behaviour model.
For example, one may assume that the current short-run behaviour of the government
by the way of monetization of fiscal deficits is the fiscal rule in place (even though this
will not be regarded as a fiscal closure rule in the sense that it does not ensure that
intertemporal budget constraints are respected). This will require the elaboration of
specification of money supply equation to capture the links between fiscal deficits and
money supply, and since money plays a role in determining price inflation, the
consequences of the monetization for the real exchange rate, interest rates, prices and the
general behaviour model can be investigated. But the model will not have a sensible
long-run solution because the fiscal rule in place is certainly not sustainable: the
government cannot forever finance its deficits by monetizing it. Of the alternative
closure rules that can be entertained for the model, the one we have adopted for the
benchmark model assumes that the government issues new debt to meet its current
deficits but constrains the Central Bank not to buy government debt. The government
is assumed to target an exogenous path of debt to GNP ratio, and therefore adjusts the
tax rate to prevent the stock of government debt from rising forever relative to GNP. In
this case, the transmission effects of government fiscal deficit will be substantially
different.
Another potentially innovative feature of the model is the treatment of expectations.
Most large-scale modelling exercises of African economies have usually imposed adaptive
expectations. This may be partly attributed to difficulties in solving forward-looking,
rational expectations models. However, it can be said that both of the currently used
methods of modelling expectations formation adaptive and model consistent expectations
are extreme characterizations and probably wrong. True behavioral relationships may
well lie somewhere in between the extremes (Bryant, 1992). Economic agents in
Nigeria, especially under the SAP regime, can be presumed to take their decisions on the
basis of both hindsight and foresight. The model is solved imposing either the adaptive
or the forward-looking expectations. Some modellers have attempted to capture the
effects of expectations by incorporating some learning behaviours but such efforts are not
without some serious criticisms. In the absence of an agreed procedure for building the
half-way house between the two extremes, we propose to present and compare the results
of the two expectations schemes.
Summary of the model structure

The model of Nigeria presented below is a medium-sized one. It is made up of 61
equations, out of which 19 are stochastic and 42 are definitional equations. There are
about 23 exogenous variables and parameters in the model. It is presented in five major
blocks, namely, domestic absorption, external trade, monetary sector, production and
prices, and external debt.
The model combines elements of the Keynesian framework (to permit evaluation of
short-term stabilization issues) and the neoclassical long-term growth properties (see Khan
and Montiel, 1989). It incorporates some issues characteristic of Nigeria's debt profile
and government policy responses to internal and external shocks, and also captures the
economy's dependence on the volatile oil sector. It is a model of a small, open
developing economy in which external and internal shocks determine macroeconomic
outcomes. The basic assumptions that underlie the modelling strategy are those that are
consistent with the policy reforms under the SAP regime, namely, open current account,
liberalized financial and trade sectors, and flexible exchange rate.
While capital mobility is assumed, it is recognized that Nigeria is a heavily indebted
developing country and there may he several factors that may make claims on it carry
some degree of default risk. So the rate of return on Nigeria's liabilities will take
account of the expected return instead of simply the contractual obligation. For example,
external investors in Nigeria will demand a premium on new debt if outstanding
obligations are selling at less than par. Thus there will be an arbitrage relationship that
sets the domestic interest rate equal to the world interest rate adjusted for default risk.
The assumed arbitrage conditions are predicated upon an assumption of capital mobility.
Traded goods prices (oil, commodity and import prices), interest rates and exchange
rate are the primary means through which external factors affect the domestic economy.
Another important channel of linkage to the rest of the world is the external debt service
payments. Debt service payments have impact primarily on the government fiscal deficit,
and through the government's borrowing and endogenous tax behaviours affect the
monetary and real sectors of the economy. Investment is derived from capital stock
adjustment framework that follows a partial adjustment mechanism, while private
consumption follows a liquidity-constrained and unconstrained inter-temporal utility
maximization. Exports and imports are modelled in the conventional manner with
relative prices and activity being the major determinants. Prices in the domestic economy
follow a Philips curve relationship that allows for overlapping contracts. Therefore both
prices and wages are sticky in the short-run- permitting us to specify short-run output
determination in a Keynesian manner while capacity output is determined by a
neoclassical production function. The debt sector is elaborately spelt out, and an attempt
is made to capture the relationships between the external debt and government's
budgetary behaviour.
As indicated earlier, the model is solved by imposing either the adaptive or model-
consistent, rational expectations in both the goods and financial markets. Forward-
looking expectations is an important innovative feature of the model results. For
example, changes in policy today - or expected changes in policy tomorrow - will have
an immediate effect on the jumping variables - the exchange rate, interest rate, and
prices. The detailed specifications of the model equations are laid out below.
Figure 1. Flow chart showing major linkages in the model blocks.

Block A model specifications: domestic absorption
Domestic absorption comprises private consumption expenditures (CP), private

investment expenditures (INV) and exogenous government expenditures (G) on goods and
services (that is, total expenditures, GEXP, less total debt service payments, TSERV).
Private consumption expenditures

A single private consumption equation is estimated without disaggregating the total
expenditures into purchases of durables and non-durables. The specification is a fairly
conventional one and follows closely the specifications by Blinder and Deaton (1985),
Lahiri (1989), and Haque, et al (1991). The equation is given as follows:
= a0 +allnYD1 +aIn4RLENR (1.1)
where ô is first difference operator, YD is real current disposable income and RLENR
is real lending interest rate (long-term rate). Real disposable income is defined as the
nominal gross domestic product less total income tax receipts (net of transfers)
(YTAX), plus government debt service payments to domestic residents (DSERV), plus
interest receipts on dollar denominated foreign assets of the private sector (PFA),
minus interest payments on total debt owed by the private sector to external private
creditors (NFPCAP), and minus interest payments on claims of the domestic banking
sector (PCR).
YD, YTAX + DSERV + (URS*PFA,,)1ER -

(URS*NFPCAP,I)/ER - (LR*PCR) )/PGNP (1.2)
PGNP is the GNP deflator; URS is the US short-term interest rate used to calculate debt
interest payments; and ER is the nominal exchange rate (naira per US dollar).
Alternatively, disposable income and consumer expenditure can be linked to the net
change in consumer wealth by the private sector budget constraint given as:
YD, = CP, -l-INV, + ((MT, +(PFA,-PFA,,) -

(NFPCAP,-NFPCAP,,) -(PCR,-PCR,1) +(DDEBT,- }/PGNP
where INV, is private investment expenditures; MT, is the target stock of money
supply; and DDEBT, is government bonds (total domestic debt of the government).
All other variables are as previously defined. The above specification implies that
disposable income is allocated to consumption, investment expenditures and net

changes in financial assets.
Private investment expenditures

Investment function is derived from a capital stock adjustment framework. The
specification follows closely from the specifications of investment functions by Ahmed
and Chhibber (1989), and Chhibber and Shafik (1990).
Nigeria depends heavily on imported capital goods and raw materials for most of its
investment expenditures. It is therefore expected that the persistent depreciation of the
local currency since the introduction of the adjustment programmes will have impacts on
the cost of imported inputs, the supply price of capital, real wages, interest rate, and
hence on the short-run profitability of investment expenditures.
Desired private capital stock (KP*) is hypothesised to be a function of expected demand
(GDPe), the rental cost of capital (domestic and world prices of capital) proxied by the
real domestic long-term interest rate (RLENR), the real cost of imported inputs
(approximated by the real exchange rate [REXR], and government capital stock (KG).
In a developing country such as Nigeria, a significant proportion of the government's
capital expenditures is invested in the provision of basic infrastructures (roads,
telephones, electricity, water supply) and these increase the productivity and profitability
of private investment. It is expected that the negative short-run effect of REXR on KP*
is due to the higher real cost of imported capital goods that are not easily substitutable
at home and the higher real cost of intermediate goods required for production, reducing
the incentive to invest. Both the rental cost of capital and the real exchange rate are
expected to negatively affect investment expenditures.
Desired capital stock equation might thus be specified in linear form as:
KP*
= +a1GDPe +a,REXR, +a4KG (2.1)
The capital stock adjusts to its desired level with an adjustment lag fi, which takes a value
between 0 and 1.
KP, = (2.2)
The evolution of the capital stock is given by
= - Dep*(KP11) (2.3)
where Dep denotes the rate of depreciation.

Substituting (2.1) and (2.3) into (2.2) gives the equation to be estimated for private
investment expenditures as:
= +
(Dep . (2.4)
An identity for the gross capital stock can be derived from equation (2.3) as follows;
KP, = +INV1 - Dep*(KPt1) (2.3a)
Government sector
In this section, only the activities of the federal government are modelled. Ideally, it
would have been better to capture the activities of all levels of government (federal, state
and local) and public enterprises, but comprehensive and up-to-date data on the activities
of other tiers of government are not easily available. In any case, given Nigeria's fiscal
structure, the federal government's fiscal activities capture, for all practical purposes,
Nigeria's public finance. The state and local governments depend on the federal
government's statutory allocation for over 80% of their revenues; all external public
debts are liabilities of the federal government and about only the federal government sells
debt instruments in the securities market to finance its deficits.
The revenue-expenditure relationships of the federal government are modelled. One
objective is to capture the implications of the persistent depreciation of the naira (local
currency) in the face of huge and rising external debt-service obligations on government's
fiscal balance. Reducing the government's fiscal deficits (expenditure-reducing) and
devaluation of domestic currency (expenditure-switching) are two major policy
instruments of the SAP package in Nigeria. Real depreciation of the naira improves the
current account, but it also increases the naira value of debt-servicing obligations and
imports, thus widening the budget deficit and stimulating the government's recourse to
domestic borrowing and inflationary finance.
Government revenue
It is assumed that all tax revenues in the economy accrue to the federal government,
which in turn shares all the collected revenues between itself and the other two tiers of
government (state and local governments). The major sources of revenue to the
government can be grouped into oil sources (accounting for over 75% of total revenue)
and non-oil sources. Oil sources include the petroleum profit tax, revenues from rent and
royalties of the oil companies and the earnings of the Nigerian National Petroleum
Corporation (NNPC) from domestic consumption of petroleum products. The non-oil
sources include company income tax (about 2% of total revenue), customs and excise
duties (10%), personal income (labour) taxes which account for about 10%, and other
revenues. All these federally collected revenues are paid into the federation account and
shared among the federal, state and local governments. Equations for the major revenue
sources are specified below.
The equation for petroleum profit tax is specified to depend on the nominal naira value
of oil export receipts.
The equation is specified as follows:
(1)
OILTAX is total nominal oil tax revenue and OILEXP is oil exports in constant
naira.
Other oil related revenues (OTHOIL) are mainly the domestic oil sales revenue for the
Nigerian National Petroleum Corporation (NNPC), and the rents and royalties paid by
the oil companies. These revenues are determined by the level of aggregate demand and
the domestic price of oil:
=a0 +ai(GDP*PGNP)t (3.2)
where DPOIL is the domestic price of oil (which is fixed by the government and thus
taken as exogenously given.
Revenue from taxes on the corporations is specified as a function of nominal income.

Non-agricultural income might seem preferable as a base on the grounds that non-
corporate farming dominates agricultural production in Nigeria. It would however be a
serious omission to ignore the relatively small but growing number of large commercial
farms that are now involved in agricultural production, especially since the introduction
of the structural adjustment programme in 1986. Revenue from corporation taxes is
therefore specified as follows:
(2) COMTAX + ai(GNP*PGNP) (3.3)
where COMTAX is company or corporation income tax.

The revenues from customs duties are defined as:
(3) CUSTAX = tarifM*IMP*PGNP + OILEXP)*PGNP (3.4)
where CUSTAX is customs and excise duties; tarifM is average tariff on imports;
tarifX is average tariff on non-oil exports; IMP and EXP are real naira values of total
imports and exports respectively; and OILEXP is total oil exports.
Following the work by Masson, et al (1990:11), nominal income tax revenue net of
transfers (TRASF) can be defined as:
= TRATE*( Dep*(PGNP*KP11) + -
TRASF; (3.5)
where YTAX is total income taxes; TRATE is the average tax rate on personal
income; Dep. is the depreciation rate on capital stock (KP), and DRS is short-term
domestic interest rate. Total income tax receipts are therefore related to a tax rate and
a tax base, which is approximated by net national product plus interest receipts on
domestic government debt. It is assumed that income taxes also accrue to the federal
government. This is also paid into the federation account and the total revenue in the
account is shared among the tiers of government.
Total nominal revenues can then be defined as:
= OILTAX + OTHOIL +COMTAX + CUSTAX + YTAX (3.6)
The federally retained revenue (GTAXF) is given as a percentage of total revenue, that
is:
GTAXF = fed*(GTAX) (3.7)
where fed is the exogenously determined share of the federal government in the
federation account.
A few observations can be made about the narrow and specialized revenue base
discussed above. Since Nigeria is both a quantity and price taker with respect to her
export of oil, changes in either of these have serious consequences on the government's
budget balance. Also, liberalization of trade that results in the lowering of import and
or export tariffs will affect the government's revenue. Diversification of the revenue
base may be difficult to accomplish in the shortrun given both the structure of the
productive sectors and the technical and administrative defects in tax assessment and
collection. It is not easy to assess and collect taxes from the agricultural and wholesale
and retail trade sectors, which are the major employers of labour. Therefore,
government's revenue-expenditure relationship is at the moment largely determined by
developments in the oil sector. It should also be noted that the flow of financial aid to
the government has been ignored in the discussion of revenue sources principally because
of its insignificance as a revenue source.
Intertemporal fiscal closure rules

There are various ways in which sets of equations representing various relationships can
be conditioned to have sensible mathematical solutions. As Harris (1985: 15) observes,
these conditions or closure rules are the most important features in determining model
response to policy interventions.
During the simulations, alternative assumptions about the government's fiscal
behaviour can be entertained. These assumptions (fiscal rules) are designed to respect
the intertemporal budget constraint, otherwise the model will not have sensible long-run
solutions. Because of the central role of fiscal policy in Nigeria's economic life, the
nature of these alternative closure rules has important implications for the modelling and
behaviour of the rest of the model, especially the monetary sector, and the nature of
linkages among the relationships modelled. Fiscal discipline is one of the goals of the
adjustment process but one that the government has had the most difficulty in attaining.
It is therefore of interest to investigate the implications for the economy of imposing
some specific fiscal rules on the government's behaviour.
Moreover, one of the two versions of the model incorporates model-consistent,
forward-looking expectations. With the programme of financial liberalization in Nigeria,
it may not be unreasonable to presume some forward-looking behaviour in the financial
asset markets. Generally, concern about intertemporal budget constraints and thus about
the sustainability of current behaviour in the medium-to long-run is an important feature
of forward-looking, rational expectations models. The likely fiscal closure rules that can
be entertained during simulations are discussed below.
The fiscal rule to be implemented in this study assumes that the government has the
power to issue new debt that is held by the domestic private sector (commercial banks
and non-bank private sector) and by external investors. This rule entertains another
assumption that is contrary to actual historical experience but consistent with the
proposals for fiscal discipline and the need to eliminate recourse to monetization of
budget deficits. Under the rule, government debt cannot be bought by the Central Bank
of Nigeria. However, the government is assumed to target an exogenous path of debt
to GNP ratio, and therefore adjusts the tax rate on personal income to prevent the stock
of government debt from rising forever relative to GNP. In other words, this rule
forecasts the changes in taxes necessary to stabilize the debt-to-GNP ratio to a
predetermined level after a given time. A feedback rule can be specified for the tax rate
that makes it respond to both the level and the change in government total debt (domestic
and external debt) relative to an exogenous target level (TDEBT.r*). Following the lead
by Mason, et. al. (1990), the difference between the actual and target level is divided by
nominal GNP, so that a given deviation in the debt stock relative to GNP will elicit an
equivalent tax rate response. This is a simulation rule with imposed parameters and can
be specified as:
= DUM* { a1((TDEBT
+
a2((ÔTDEBT- oTDEBTF*)11/(GNP*PGNP)) } (3.8)
where ô is the first difference operator; DUM is an imposed dummy variable used to
adjust the tax rate reaction function; TDEBT is total debt stock (domestic and
external); TDEBTr* is target debt stock; and a1's represent the imposed parameters.
This closure rule has important implications for the government's fiscal behaviour and
also for the behaviour of the model. Since the targeted debt-to-GNP ratio includes the
external debt component, it implies that any exogenous shock that raises the debt service
payments and therefore either raises total government expenditures in the present period
or raises the amount of unpaid arrears will result in the adjustment of the tax rate
immediately or in the future. This will be true if the targeted debt-to-GNP ratio is to be
attained and then maintained. So far, the focus of this tax rate reaction function is the
tax rate on personal income. This is maintained for the analytical benchmark model, but
it is possible to specify the reaction function such that revenue sources other than the
income tax are varied. For example, the government could raise the tariffs on external
trade and! or the domestic price of oil. Any of these will lead to increased revenues to
the government, but may also undermine some other aspects of its adjustment process.
The crucial point is that whatever rule is enforced should be such that the intertemporal
budget constraint will be respected. The rule could also illuminate policy analysis by
forecasting the amount of domestic resources that must be mobilized if a particular
expenditure profile is to be sustainable.
The previous tax-rate assumption about the fiscal behaviour of the government is only
one among many assumptions that can be entertained during the simulations. However,
the nature of any particular fiscal closure rule also determines the nature of the
assumptions that can be entertained in modelling the monetary sector, and thus the
relationship between the monetary and fiscal policy. For example, an alternative fiscal
rule that could be specified would constrain the government not to issue additional debt,
but to adjust taxes so as to keep budget balance. Nigeria's total debt-to-GDP ratio is
already very high and some analysts may argue that a sensible and sustainable fiscal
behaviour would require the government not to issue new debts but should adjust its taxes
so as to meet its expenditures (including debt service payments). A tax rate reaction
function which adjusts tax revenues to maintain budget balance can thus be specified.
Any increase in government expenditure in excess of current revenues would then have
to be met by collection of additional tax revenues. An important practical question,
however, relates to the feasibility of increasing non-oil tax revenues in the short to
medium term. This rule may raise the need for creative re-thinking of Nigeria's fiscal
federalism (in terms of flexibility of shares of the federation account in the face of
external debt service payments jurisdiction over certain revenue sources additional
revenue sources, improved efficiency in taxing the under-taxed tax base, etc).
Alternatively, the fiscal constraint may inject the needed discipline into the government
expenditure profile.
Another fiscal rule similar to the one above may constrain the government not to issue
new debts and not to raise taxes, but to meet all its current expenditures (including
changes in debt service payments) from its current revenues. This rule is akin to the
balanced budget requirement and some analysts argue that it is one of the more effective
ways to enforce fiscal discipline. If the government were to meet all its debt service
payments, then the component of government expenditures devoted to the purchase of
goods and services would have to adjust endogenously to available tax revenues. In other
words, the government is constrained not to have an independent fiscal policy. Given
the already very large stock of government debt (and thus the high debt service
payments), this rule would mean that in the face of declining revenues or an exogenous
shock that increases debt service payments, most of the total expenditure outlays would
have to be devoted to debt service payments. The implications of this fiscal rule for the
behaviour of the model will be interesting to examine.
Still another fiscal rule could mimic the current behaviour of the government, which
may not be sustainable in the long-run. Under this assumption, the government issues
new debt to cover any shortfall of revenue relative to outlays. This may be caused by,
among other things, an increase in the external debt service payments or a fall in
revenues. Most of the debt instruments are sold in the domestic market and are bought
by the Central Bank, commercial banks and the non-bank private sector. Experience has
shown that sometimes a major part of the debt instruments is held by the Central Bank
of Nigeria. Domestic component of the government's total debts has therefore been
rising explosively. In this regime, the government does not have an independent
monetary policy, and an endogenously determined monetary policy will be modelled.
However, this assumption cannot be regarded as a stable fiscal closure rule, because it
does not ensure that the government respect the intertemporal budget constraint. A
version of the model incorporating this behaviour may be created to study the
implications of this apparently unsustainable budget behaviour on the behaviour of the
model.
Government expenditure
The non-discretionary part of government expenditures (essentially the debt service
payments) is modelled, and the rest is taken to be an exogenously determined policy
variable.
In spite of the various efforts (at least during the budget preparation) to ensure
balanced budget or even provide for moderate surplus, exploding budget deficits, ex post,
have been a recurrent phenomenon especially since the adjustment programme. Given the
emphasis of the adjustment programme on fiscal discipline, this phenomenon would
appear to be paradoxical and is typical of the experiences of some other developing
countries (especially in Latin America) undergoing the adjustment process. Efforts have
been made to privatize and commercialize several of federal government's parastatals and
also to review or remove some government subsidies. But it has been difficult for the
government to contain the deficits. Depreciation of the local currency ensures that
government realises increased naira revenues from oil exports. At the same time, the
persistent depreciation of the naira and rising domestic inflation also lower the real value
of government expenditures and thus necessitate an increase in the naira content of
budgeted government expenditures so as to maintain the same or even lower level of real
spending.
Beside the impacts of currency depreciation and political considerations that constrain
the government from drastically cutting expenditures in several sectors (such as defense),
there is one other principal factor that reinforces the deficits. This factor is the increases
in domestic and external debt service payments. Over the period 1988 - 1990, debt
service payments as a percentage of total expenditures were 40.7, 49, and 67
respectively. In 1990 for instance, total debt service payments increased by 104% over
the 1989 level, and government expenditures as percentage of GDP correspondingly
increased, from about 21 % in 1989 to 26% in 1990. Also in 1990, the stock of domestic
debt increased by 47.4%, while that of external debt (in naira) rose by 24%. The
increase in external debt was attributed to "draw down of external loans, capitalized
unpaid interest charges on rescheduled debt obligations, depreciation of the US dollar
against other currencies and depreciation of the naira exchange rate" (see Central Bank
of Nigeria, 1990:60). As the bulk of Nigeria's huge external debt is owed to private
creditors at commercial rates, it is expected that variations in the world interest rate
would affect the debt interest payments. Given the rate of growth of debt stock and the
size of debt service payments as percentage of total expenditures, it is logical to expect
that without drastic reductions in the other expenditures, budget deficits will be difficult
to eliminate in the medium term.
As indicated earlier, the deficits are financed through borrowing. Since the existing
debt-burden is high, it may be expected that potential bond buyers would attach a high
default risk to government bonds and thus be reluctant to buy them. If, in addition, it
is assumed that capital mobility is not perfect (which is not the assumption entertained
in the benchmark model), then the low domestic demand for bonds will drive real interest
rates above world levels. The ratio of domestic debt to GDP rises. If a bulk of the
government's debt instruments are purchased by the Central Bank of Nigeria, however,
then money creation plays an important part in the govermnent's efforts to make the
internal resource transfer required to service its debt obligations, with enormous
consequences on the economy's broad macroeconomic balance. Our benchmark model
assumes that the new debt issues are not held by the Central Bank but by the domestic
private sector and external investors. The specifications that follow are based on this
assumption, but alternative versions of the model consistent with the various closure rules
discussed earlier can be created during policy simulations.
Government's total nominal expenditures (GEXP) are divided into two: the exogenous
expenditure on goods and services (G), and debt service payments (TSERV). The total
debt service payments are then divided into the payments to service domestic debt
(DSERV) and the actual payments made to service external debt (VOLSEV)3. These
service payments represent an important linkage between changes in government's debt
service obligations and the budget balance. In 1990 for instance, these payments
constituted about 67% of total government expenditure, and correspondingly, domestic
debt rose by 47%, external debt by 24%, and Ml increased by 45%.
The major identities are as follows: Total nominal government expenditures (GEXP)
are given as the exogenous expenditures plus total debt service payments;
GEXP, =(PGNP*G, +TSERV (3.9)
The total debt service payments are given as:
TSERV, =(DSERV, (3.10)
and the government's nominal current budget deficit (GDEF) is given as the total nominal
expenditures minus the total retained revenue of the federal government.
GDEF, = GEXP, -GTAXF, (3.11)
It is assumed that the government fully services the stock of outstanding domestic
debt. Each period, a proportion of the domestic debt is due for repayment, and the
government retires this amount and also pays interest on the outstanding obligations. For
the purposes of this study, we ignore for the moment the gross amortization and gross
new issues, and focus on net change in the stock of debt. Domestic debt service
payments are therefore specified as follows:
DSERV, = (3.12)
where DDEBT is the stock of domestic debt and DRS is the domestic short-term
interest rate.
New issues of domestic debt add to the stock. The government is assumed to fully
service its domestic debt, and changes in the stock of domestic debt equal the new issues
of domestic debt:
= + Dom* GDEF1 (3.13)
where Dom is the proportion of government current deficits financed by borrowing

in the domestic market, and is taken to be exogenously determined.
The above specifications of the government revenue-expenditure relationships imply

that changes in external factors (oil price, OPEC oil production quota for Nigeria,
external and domestic debt service obligations) have serious implications for the
government's budget behaviour.
Block B model specifications: External sector
Exports
Exports are disaggregated into three goods: oil, primary commodities (mainly agricultural
commodities) and manufactures. Oil exports account for over 90% of Nigeria's total
export receipts. Total oil export at given international price of oil is determined by
OPEC production quotas. At a later stage of the research when the model would be
simulated interactively with some global models, the dynamic path of oil price can be
derived from the global models. For example, in the IMF's global model, the
MULTIMOD, the nominal price of oil is determined by the interaction of world
aggregate supply and demand. Each region's oil consumption is a function of the
region's real GDP and the price of oil relative to that of aggregate output. Imports of
oil by the region (except the high-income oil exporting countries) are the residual
between net domestic supply and consumption. The net oil exports of oil exporting
countries then equal the sum of the excess demands of the industrial countries and non-oil
exporting developing countries.
The volume of Nigeria's total oil export receipts (in constant units of local currency)
for each year can be defined as follows:
OILEXP, = (4.1)
where OILEXP is total real oil exports; PQOTA is the proportion of OPEC production
quota for Nigeria that is exported; POlL is export price of oil in US dollars; and ER is
nominal official exchange rate.
Manufactures export is not as yet a significant proportion of Nigeria's total exports
(less than 2%), but this classification is maintained to account for the export of
manufactures and semi-manufactures of agricultural products, textiles, and every other
export item not classified as oil or primary commodities. As a group, these export items
constitute a very insignificant proportion of world manufactures exports. Their exports
are assumed to be constrained by domestic supply conditions. Changes in exports might
be specified to be constrained by changes in the total output of the non-oil sectors of the
economy as the scale variable. Real exchange rate (REXR) is included to capture the
effects of relative prices, and especially the impact of currency depreciation, which also
affects the competitiveness of this class of exports in the world market. The average
price of developing region's manufactures exports (PXM) in MULTIMOD is also used
as a proxy for the prices facing Nigerian manufactures exporters. We have not adjusted
the export price for export taxes because export duties are generally negligible in Nigeria.
For example, there is no export duty on oil exports, while duties on non-oil exports have
been very insignificant and indeed have been abolished since the adjustment process
began in 1986. Since the specification is an export supply equation, the price variables -
REXR and PXM - are expected to have positive coefficients. The equation for export
of manufactures is therefore given as:
ÔInMAEXP, = a0 (4.2)
where ô is first difference operator, 0 is the share of non-oil sectors in the GDP and
MAEXP is the real manufactures exports.
Export of primary commodities (agricultural) (PMEXP) mainly follows the pattern

of the specification for the manufactures exports. However, instead of the export price
of manufactured exports, the naira value of the exogenously determined world price of
commodities (PCOM*ER) is included as the relevant price variable. Exporters of
primary commodities in Nigeria are assumed to be price- takers in the international
commodity market. Nigeria is a negligible supplier to the world commodity market and
an export supply equation for primary commodities is specified. Exporters are assumed
to be influenced by the relative prices in their supply decisions. The equation for real
export of primary commodities is given below with all the variables expected to have
positively signed coefficients:
ÔInPMEXP, = a0 + + (4.3)
ADJUSTMENT, TRADE AND GROWTH 31
Total real exports (EXP) is therefore given as;
EXP, =(OILEXP, +MAEXP, (4.4)
An aggregate exports price can also be derived given that:
Oil export price = POlL/ER

Primary commodity export price = PCOM/ER
while the price of manufactures exports =PXM/ER;
then total exports price is given as:
(OILEXP*POIL/ER + PMEXP*PCOM/ER +
MAEXP*PXM/ER)/EXP (4.5)
Imports
Imports in Nigeria can be divided into two major activity groups: consumer goods, and
capital goods and raw materials. Estimating separate equations for raw materials and
capital goods reflects the fact that local manufacturing firms depend heavily on imported
capital goods, and that the utilization rate of domestic industrial capacities depends on the
adequacy of imported raw materials. Import of raw materials and capital goods (CAIMP)
is assumed to be related to the production level and the capacity utilization rate in the
manufacturing sector. Import of consumer goods (CNIMP) is a function of real
disposable income as the relevant activity variable. In the two import functions, a
measure of the real cost of imports (REXR), that is the real exchange rate, and real
lending rate (RLENR) are included. Lagged dependent variables are also included to
capture partial stock adjustment behaviour. Real exchange rate is defined later as the
ratio of import prices (in local currency, DPIT) to domestic prices. The import equations
are thus specified as follows:
Ô1nCAIMP, = a0 +a31nRLENR1
(4.6)
=a0 + (4.7)
Total imports (IMP) can then be defined as
IMP = CAIMP +CNIMP (4.8)

The price of imports in naira is derived as follows:
DPIT, = ((IMP*DPIM)/ER)/IMP (4.9)
Trade balance (in local currency) can thus be defined as the difference between total
exports and total imports:
TBAL, =(DPXT*EXP (4.10)
Block C model specifications: Monetary sector
In this sector, the equations specified include those determining the demand and, or
supply of money stock, interest rates (short and long term rates) and the exchange rates.
Alternative assumptions about the government's intertemporal budget constraints may be
crucial in determining which of the monetary variables can be exogenized or
endogenized. For example, if it is assumed that the government faces an exogenous
budget constraint and so cannot issue new debts but must meet all its expenditure and
debt service payments either from current revenues or through changes in taxes, or that
the new debt issues are not bought by the central bank, then it may be realistic to
experiment with versions of the model alternatively endogenizing or exogenizing money
supply. Under the assumption that the government debt instruments are held by the
Central Bank (and this represents the current practice in Nigeria), then money supply can
be treated as endogenous. The equations specified below are those that are relevant for
the benchmark model. Appropriate modifications in the equations consistent with the
various fiscal closure rules can be made during policy simulations.
Equations specified are those of money demand, a central bank's reaction function
determining short-term nominal interest rate, long-term nominal interest rate, real short-
and long-term rates and exchange rates (nominal and real). Disequilibrium effects in the
money market affect primarily the interest rate and prices, and through these channels
affect employment and output, financial flows, etc. The major channel through which
monetary policy instruments affect the domestic economy is the cost-of-capital effect.
Monetary policy is assumed to affect directly the short-term interest rate and indirectly
through a term-structure-of-interest rate relationship to affect the long-term rate. The
interest rate variable appears in the investment equation, thus permitting the IS-LM type
linkage between the financial and real sectors.
Demand for real money balances is specified to be a function of real income (GNP),
nominal short-term interest rate and the current and expected inflation rates. It is
conventional to specify money demand function to depend on income and nominal interest
rate, where the nominal interest rate measures the opportunity cost of holding money.
Inflationary expectation is also specified as an alternative measure of the opportunity cost

of holding money in developing countries. It is argued that because of the
underdeveloped financial markets in developing countries and the high inflationary levels,
agents try to hedge against inflation by investing in physical assets rather than holding
money. The inflationary term is therefore expected to be negatively related to the
demand for money. Our specification below includes both the nominal interest rate and
inflation terms in the demand for money function. However, estimation results indicate
that the interest rate variable is a more appropriate measure of cost of holding money in
Nigeria. A lagged dependent variable is also included to capture the partial stock
adjustment process in the demand for money function. The equation is given as:
= a0 +
a5lnINFe (5.1)
where M is the desired money stock, DRS is nominal short-term interest rate, INF
is the inflation rate, and INFC is the expected inflation rate.
Interest rate
Equations for the nominal and real short and long interest rates are specified. The
equation for the short-term nominal interest rate is specified as the central bank's reaction
function. It is assumed that the central bank adjusts the short-term rate to close the gap
between a target for money stock and its actual value. Government's borrowing activities
in the domestic financial market are expected to raise domestic interest rates. This
reaction function does not necessarily capture the actual historical experience. In the
past, though the central bank has varied short-term and inter-bank lending interest rates
to elicit certain reactions from the financial system, it relied mainly on direct controls of
credit for its monetary policy. The rationale for the specification in our benchmark
model stems from the policy directions under the ongoing liberalization of the financial
system and the switch from direct to indirect instruments of monetary policy. It is
expected that the central bank will increasingly rely on the manipulation of interest rates
to control money supply. As a variant of similar specification in Masson, et al (1990),
the function is given as:
= +a2ln(MT/M)/(m3) (5.2)
where MT is the target money stock, and m3 is the coefficient of the DRS term in the
demand for money function.
The real short-term interest rate (RSR) is defined as the nominal rate adjusted for
growth in the price level in the current period:
= (5.3)
Under the consistent expectations, the long-term interest rate (LR) is related to the
short-term rate by a term-structure relationship. In the version with adaptive
expectations, the long-term rate adjusts gradually to observed innovations in the short-
term rate4. The equations are given as follows:
Consistent expectation: = (5.4)
Adaptive expectation: =
a are imposed parameters and are homogenous of degree one.
The real long-term rates (RLENR) (under consistent and adaptive expectations) are
similar to equations (5.4) and (5.4a) above but the nominal short-term rate is replaced
by the real short-term rate (RSR):
Consistent expectation: RLENR, = E70 (5.5)
Adaptive expectation: = LR/100 - INFEA (5.5a)
where INFEA is inflationary expectations when expectations are adaptive
Exchange Rate
The strategy for modelling an exchange rate depends crucially on the assumptions made
about the maturity of the financial system and nature of capital mobility in Nigeria. If
capital assets are assumed to be perfectly mobile, and hence the domestic financial assets
are perfect substitutes for international assets, then the naira exchange rate (defined as
the amount of foreign currency [US$1 per unit of local currency) can be determined by
an interest parity condition that adjusts the naira value to, say, the US dollar.
In Nigeria, however, the process of financial liberalization has led to an evolution in
interest and exchange rates determination from a system of statutorily determined rates
to gradual dc-control of the system. Currently, the rates are allowed to be market
determined. Also, it is assumed that capital is mobile. This may not be an unreasonable
assumption. In the recent past, capital controls have lessened and agents now allocate
some of their portfolios between domestic and foreign financial assets. For example, it
is not uncommon to find that individuals and firms in Nigeria hold their bank deposits
in both the local currency (naira) and US dollars. Deposits in dollars earn interest based
on the rate prevailing in the United States.
Exchange rate behaviour is therefore hypothesized to follow an interest parity
condition. The current value of the exchange rate adjusts relative to its expected future
value and default risk premium on Nigeria's external debt to equalize the domestic and
US short-term interest rates.
ER, = { (1 +DRS/100)/( (1 +URS/100)*(1 +DRP)) (5.6)
where ER is current nominal exchange rate; URS is the United States' nominal short-
term interest rate; DRP is default risk premium; and is the expected exchange
rate.
As an alternative to the above uncovered interest parity equation, the naira could be
assumed to be fixed to, say, the US dollars. The fixed rate could be with finite bands -
the so-called 'exchange rate target zones' and perhaps of the type practiced within the
exchange rate mechanism of the European Monetary System. In this circumstance, it
could be assumed that the central parity for the US dollar/naira exchange rate is set
exogenously by the Central Bank of Nigeria, but with occasional discrete changes in the
parity. For such an assumption, a different path for the central bank's reaction function
must be specified. For example, instead of the assumption that the monetary authority
moves the short-term interest rate to narrow the discrepancy between a target for the
stock of money and its actual value, we might assume that central bank adjusts the short-
term rate to maintain a given dollar/naira parity. The change in short-term interest rate
could be specified to be a nonlinear function of the gap between the exchange rate versus
the dollar and an exogenous target level5. The money supply would then be completely
endogenous.
Under the forward-looking expectations, agents are assumed to forecast the future
exchange rate accurately and thus, the expected exchange rate (ERa) can be defined as
the next period exchange rate, that is:
Consistent expectations: = (5.7)
Adaptive expectations: EREA = + a1(ER-
The real exchange rate (REXR), defined as the ratio of tradeable and non-tradeable goods
prices, is here given as the ratio of export prices (in local currency) to the domestic
price level (PGNP):
REXR, = DPXT/PGNP (5.8)

Block D model specifications: Production and prices
Production
In this section, aggregate output (GDP) and income (GNP) variables are defined. The
equation determining capacity output is also specified. It is assumed that in the short-run,
output is determined in a Keynesian manner by the components of aggregate demand, but
the long-run capacity output is determined by a neoclassical production function.
This specification may imply that the observed capacity under-utilization in the
economy results from lack of demand. Some scholars would argue however, that
because of foreign exchange constraints and other factors, short-run output in some
developing countries is supply-constrained. It is our view that the emphasis of either the
demand or the supply side is an empirical issue. The two may perhaps be extreme
characterisations of the underlying process and the choice of one to the neglect of the
other may not accurately describe the entire reality. For example, the emphasis of the
supply side would analytically imply that a version of Say's law prevails in the economy
and therefore prices are flexible enough to clear the markets. In Nigeria, for instance,
this assumption would neglect the often reported large inventory of unsold goods by
firms, or the sluggish adjustment of labour wage rate to observed innovations in price
inflation. Besides, it may be difficult to argue a case for foreign exchange constraints
in a regime of flexible exchange rate, as approximated by the current exchange rate
determination process in Nigeria.
Furthermore, recent studies of the economic performance in Nigeria under the
adjustment programme provide evidence that weak consumer demand and rising stocks
have been constraining the manufacturing sector from meeting expected growth rates.
For example, Nigeria's First National Rolling Plan 1990-1992 (page 84) observes that
some of the factors that inhibit domestic output are prohibitive cost of replacing
equipments, increase in the cost of raw materials and high interest rates. The later, in
turn, led to high cost of production of finished goods. This, accompanied by consumer
resistance, led to reduced demand.
This finding for Nigeria may not be inconsistent with what can reasonably be
expected under a SAP regime. Observed under-utilization may be symptomatic of the
radical departure under SAP, from the previously expected path of economic policies.
With the existing capacity, the new policy reforms render some past investment decisions
unprofitable due to valuations in relative prices. Relative to the level of aggregate
demand, unprofitable firms may be forced into obsolescence or operate at less than full
capacity due mainly to their cost ineffectiveness. This is the phenomenon reported for
the former communist East European countries and may also characterize the transitional
phase for many structurally adjusting developing countries.
However, whereas we recognize the relative merits of the arguments for emphasizing
one side or the other, especially the need to emphasize the supply side for some
developing countries, care should also be taken of some subtle issues that might
undermine the analytical interpretation of the results.
In any case, it doesn't seem as if the dichotomy between the supply-siders and the
Keynesians matters greatly. In a recent study, Friedman, (1991: 1) has shown that,
The distinction between aggregate demand and aggregate supply as

the principal location of the disturbances that drive business cycles -
the distinction most popularly associated with real business cycle
models -is less important from a policy perspective than is commonly
believed. The policy prescriptions that follow from these models have
more to do with the kinds of assumptions that they incorporate about
how the markets function than with whether the chief economic
disturbances work through demand or supply.
Our strategy is to model this sector in a manner consistent with the overall model
framework, to permit investigation of short-run stabilization issues as well as long-run
growth (structural adjustment) issues. While short-run output is determined in a
Keynesian manner, care is taken to model one component of aggregate demand, i.e
investment,to capture some short-run supply constraints, such as the price of imported
inputs. Furthermore, the model ensures that long-run factors (supply factors) determine
capacity (or potential) output, which in turn is used to derive a measure of capacity
utilization or demand pressure (expressed as a ratio of actual to capacity output) in the
wage and price equations, thus ensuring a dynamic interaction between short-and long-
run factors in output determination. The equations are laid out below:
Real GDP is defined as:
GDP, = CP, +INV, + G, + (EXP, (6.1)
and the real Gross National Product (GNP) is defined as:
GNP, = GDP, (6.2)
where AUR is average US interest rates on the two sources of external debt (that is,
URS and FUR) and NFA is net foreign assets in local currency.
Aggregate supply equation determining capacity output is assumed to be determined

by a Cobb-Douglas production function of the simple form:
YCAP = f(KP, KG, LF, T) (6.3)
where YCAP is capacity output; KP is gross private capital stock; KG is public

capital stock; LF is labour force; and T is a time trend that captures the effects of
technical change. Real potential output is thus specified as:
= a0 + a2KG11 + +a3T (6.4)
where all variables are as previously defined.
Employment
Employment in the model is assumed to be demand determined. This is not unreasonable
in view of the high level of unemployment in the economy. In modelling this section,
we abstract from the dualistic nature of the labour market (that is, an under-utilized
labour in the rural sector and an urban wage differential). It should be noted, however,
that the share of agriculture in the GDP and the share of the labour force engaged in it
have declined steadily since the 1960s. The rural-urban wage differential is reported to
have narrowed significantly since the adjustment programme. Besides, the supply sector
is modelled in a highly aggregative form but in a future extension of the model with
greater sectoral disaggregation, it may be necessary to experiment with different labour
market behaviours. Market forces have been playing an important role in wage and
employment determination. We have therefore assumed a homogenous labour market.
Changes in labour supply are assumed to be exogenously given and to grow at a

constant rate. Change in labour supply in the base year is measured by the number of
graduates from the educational system (primary through university levels). Equations for
labour demand and the wage rate are specified below.
Change in demand for labour is specified simply as a function of real wage and
capacity utilization rate. Capacity utilization rate (CAPUTL) is defined as the ratio of
real GDP to real potential GDP (YCAP). The equation is given as:
= a0 +a1ôlnWG, + (6.5)
where LF is the labour force employed, and WG is the real wage rate.
CAPUTL, = GDP/YCAP (6.6)
Wage determination follows the conventional Philips curve specification, with the rate
of unemployment (UE) and the rate of inflation as explanatory variables. Perfect
indexation of wage contracts to the rate of inflation is not the practice in Nigeria.
However, it is expected that wage negotiations take account of inflationary pressures.
Following the work of Chadha, Masson and Meredith (1991), the wage rate equation is
specified to capture a regime of overlapping wage contracts in which backward-looking
and forward-looking elements determine current wage rate. Past and expected inflation
rates are therefore included in the wage rate determination. The effect of labour market
disequilibrium is captured by the current unemployment rate (UE). The wage rate
equation is thus given as:
= +a2INP + + a4WGt-1 (6.6)
The excess demand term (UE) further serves to capture the neoclassical market
adjustment mechanism. During the simulations, it may be necessary to set the
inflationary feedback into wage adjustment so as to preserve price neutrality in the model
over the long run. For example, the coefficients of the lagged and expected inflation
terms may be constrained to sum to unity so that a purely nominal shock in the model
(a doubling of money suppy for example) would not have real effect on the domestic
economy in the long run.
UE, = LF-LF8 (6.7)
where LFS is the exogenously determined labour supply.
Prices
In this section, equations determining the domestic price level (in this case the GNP
deflator) and the growth rate of the deflator (price inflation, INF) are specified. Further
extensions of the model in the future will differentiate between output and consumer
prices and incorporate other prices such as the consumer price index and sectoral output
prices. The current study is highly aggregative.
The domestic price level (PGNP) is defined as:
PGNP, = + (6.8)
where INF (inflation rate) is the rate of change in PGNP.
Movements in prices are assumed to be determined as a mark-up over unit costs in

the long run. Cost elements would include smoothed changes in labour costs, capital
costs and import costs. Labour costs are measured by the wage rate (WG) while the cost
of imports is measured by the import prices in local currency (DPIT). Profit margins are
assumed to depend positively on the rate of capacity utilization (CAPUTL). A lagged
inflation rate is included to capture the partial stock adjustment, and also measures
inflationary stickiness. The inflation equation to be estimated (with all variables in logs)
is thus given as:
INF is inflation rate.
Block E model specificaions: Measuring the effect of

debt overhang
There has been growing attempt to capture the relationship between a country's external
debt and its domestic economic conditions. The linkage mechanisms between the two
would depend mostly on the nature of intertemporal budget constraint imposed on the
government budget behaviour. As discussed earlier, it is assumed that the government
issues new debts to meet any shortfall between revenues and outlays but must raise taxes
so as to respect the intertemporal budget constraint. Linkages between external debt and
domestic economic conditions are modelled to operate primarily through the government
budgetary behaviour.
A potential source of changes in debt interest payments is the change in the world
interest rates induced mainly by monetary and fiscal policies in the industrial countries.
If there are no changes in the government's oil revenues, then any such variations in debt
obligations, say an increase, would result in one of the following actions by the
government. First, it could lead to a shift of resources away from the component of its
revenues earmarked for purchases of domestic goods and services into the proportion
devoted for debt servicing. Since government spending is a prime mover of domestic
economic activities, this option would be growth-reducing. Alternatively, the increases
in debt service obligations could be translated into increased budget deficits, financed by
borrowing from both the domestic and external sources. This would have implications
for domestic money supply, interest rate, prices, exchange rate, rising stock of debt, and
private investment. It would also affect the confidence of the external creditors on the
prospects of economic recovery. A third scenario would be for the government to
reschedule or capitalize the increases in external interest payments, thereby increasing the
stock of external debt. This would affect the creditors' perceptions of the country's
creditworthiness, which might lead to capital flight and thus affect investment, exchange
rate, money supply and others. It should be noted that another external shock, say a
decrease in either or both the oil price and Nigeria's OPEC production quota, could
confront the economy with a similar dilemma. The structure of the debt sector outlined
below tries to capture these essential linkages.
Total debt shock

Equation 7.1 below defines the total stock of external debt outstanding, measured in local
currency (EDEBT), as the sum of debt owed to private creditors (PDEBT) at commercial
interest rate, plus the stock of debt owed to official agencies (ODEBT) (such as foreign
governments and multilateral agencies like the World Bank, IMF, etc.) at fixed,
concessionary interest rates. Interest rates on private debt float with movements in a key
market rate, for example, the London Interbank offer rate (LIBOR) or the US prime rate.
Total external debt is given as:
= (PDEBT1 (7.1)
Government's total debt stock, TDEBT, (domestic and external) is thus defined as:
TDEBT = EDEBT +DDEBT (7.2)
Debt service payments comprise amortization and interest payments. However,

service payments are modelled only in terms of interest payments. Contractual debt
service obligation on each component of debt stock is therefore defined as the interest
payments due on the particular debt. Contractual debt service obligation on private debt
is thus given as:
= (7.3)
where URS is the US short-term interest rate used for calculating debt interest
payments, and PDTSEV is the contractual interest payments on private external debt.
Note, however, that the debt variables (PDEBT and ODEBT) are converted from
their dollar values to naira (using the exchange rate).
Similarly, contractual debt service payments on the official debt are given as:
= FUR*(ODEBTII) (7.4)
where FUR is the fixed US interest rate charged on the debt stock, and ODTSEV is the
contractual interest payments on official foreign debt. Total contractual debt service
obligation on external debt is therefore equal to the sum of (7.2) and (7.3) above:
TDTSEV, =PDTSEV + ODTSEV (7.5)
It is assumed, however, that the government does not necessarily meet its contractual
debt service obligations period by period. Depending on its economic circumstances, it
may decide to reschedule some of the repayments and capitalize some or all of the
arrears, or, in unusual cases of unexpected boom in economic activities, it may
decide to retire some of the debts before they are due. It is assumed that the government
always tries to service a proportion of its total contractual obligations, but it also
increases all servicing as its revenues (total oil export receipts) increase. This actual or
voluntary debt service behaviour may be thought of as the government's debt service
reaction function (VOLSEV), and is specified thus:
ÔVOLSEV, = a0 +a2ÔOILEXP (7.6)
where ô is first difference operator, and other variables are as previously defined.
The difference between the voluntary debt service payments and the contractual
obligations may be thought of as involuntary lending, which is defined here as the
accumulated arrears on debt service obligations (DARREA):
= TDTSEV, -VOLSEV, (7.7)
Changes in the private and official debt stocks can now be defined. Stock of debt can
increase as a consequence of accumulated arrears and new issues. The current stock of
private or official debt is defined as the stock of the previous period's stock of debt plus
the proportion of debt service arrears accumulated on the debt source, plus any new
issues. The accumulation of arrears a particular debt stock, say private debt, is
assumed to be proportional to the size of the debt source relative to the total debt stock.
Changes in private debt stock, for example, are defined as:
PDEBT, = +jz (DARREA) + NPDEBT (7.8)
where is the proportion of arrears owed to the foreign private creditors and
NPDEBT is new debt borrowed from private foreign creditors by the government.
Similarly, the stock of official debt is defined as:
ODEBT, = + €(DARREA) + NODEBT (7.9)

where is the share of total arrears owed to official creditors and NODEBT is new
debt owed to official creditors. The sum of and equals unity.
Equations for the flow of new external private debt can now be specified. It is
assumed that the inflow (net lending) or outflow (capital flight) of private capital depends
on the private agents' assessment of the domestic economic conditions. It is hypothesized
that the government's debt service reaction function is the indicator of the economy's
overall creditworthiness and sound economic management. For example, if the
government is rescheduling its debts and capitalizing interest arrears, private agents may
fear that the actions may not be sustainable and the government may resort to any
measure to service the debts in the future, including high taxation on capital, large
devaluation of the local currency, etc. It may be expected then that private capital will
tend not to flow in, but may even flow out of the economy. The flow of new loans to
the domestic private sector therefore depends on government's debt servicing behaviour.
It is assumed that the domestic private sector fully services its stock of external debt.
External agents still attach a risk factor to their lending to the private sector because of
the economy's high debt burden. The private to private capital flow is assumed to
depend on the changes in the volume of non-oil exports, as well as government's
voluntary debt service performance.
= a0 + (7.11)
One probable interpretation of Equation 7.11 above is that even though capital is mobile,
the economy faces an endogenous supply schedule for external financial resources. This
supply depends on a forward looking assessment of Nigeria's debt-servicing capacity.
The flow of new loans from private external lenders to the government is a residual.
The government's borrowing requirement (budget deficit) is assumed to be financed by
borrowing either from the domestic market or abroad. But there are two kinds of
external lenders the official creditors and private lenders. It is assumed that whatever the
government is unable to borrow from the domestic market and official creditors is made
up by borrowing from private lenders. New external debt to private foreign lenders is
therefore given as:
NPDEBT = GDEF - (NODEBT + (GDEF*Dom)) 7.12
where NODEBT is flow of new official loans to the government, which is

exogenously determined; GDEF is government's nominal deficit; and Dom is the
proportion of deficits financed by borrowing in the domestic market.
We can now specify the equations for current account balances and net foreign assets.
Current account balances are defined as the trade balance plus interest earnings on net
foreign assets:
CURBAL, = TBAL, (7.13)
where AUR is the average interest rate on the two sources of external debt, that is,
URS + FUR = AUR.
AUR = (URS + FUR)/2 (7.14)
Net foreign assets in turn is defined as:
NFA1 = TASETS -EDEBT (7.15)
where TASETS is the total external assets of the economic agents in the economy
(official and private), and other variables are as previously defined. The total
external assets are defined as the sum of foreign assets owned by the private sector
(PFA) and those held by the government (GFA).
TASETS = GFA + PFA (7.16)
Default risk premium

It is assumed that the default risk premium does not follow a simple linear form. Risk
premium on the external debt is hypothesized to depend on the multiplicand of the ratio
of actual to contractual debt interest payments on external debt and the ratio of external
debt to exports.
)5
DRP, =( (VOLSEV/TDTSEV)*(EDEBT/EXP*PGNP) )*(AUR/100) (7.17)
where DRP is default risk premium, AUR is average interest rate for calculating
interest payments on external debt.
Estimation results
Annual data for the period 1970-1991 are used in the estimation. The data series for the
estimation span the period before 1986, that is, before the commencement of the
structural adjustment programme. Given the supposedly significant structural changes
after 1986, the stability of the time series data and the model results are of major
concern. Besides, due to the non-stationarity of most macroeconomic time series, it is
very important to avoid the problem of spurious correlation, or inconsistent regression
that plagues econometric estimation when some or all of the individual series are non-
stationary (see Adam, 1992; Nelson and Plosser, 1982; and Granger and Newbold,
1974).
Our major challenge in estimating the equations of the model, therefore, is to obtain
a sensibly congruent model. By this, we mean a model that is interpretable in terms of
the economic theoretic relationships being modelled and which also describes the short-
run and long-run characteristics of the data in a statistically robust manner.
The theory of cointegration provides a powerful, formal framework for capturing the
long-run relationship between non-stationary series (if such equilibrium relationship
exists) within a stationary model (see Engle and Granger 1987). It is therefore a method
of avoiding both the spurious and inconsistent regression problems that would otherwise
occur with a regression of non-stationary data series. Cointegration is a long-run (static)
specification theory, but it is also shown to be consistent with the error-correction
dynamic specification (Engle and Granger, 1987). It can be a particularly useful
approach in unrestricted (non-normalized) equations that are consistent with long-run
equilibrium but may be characterized by considerable short-run dynamics. The reason
for this is that the error-correction models capture the time series properties of variables,
through the more robust dynamic structure allowed, whilst at the same time incorporating
an equilibrium economic theory (Elbadawi and Schinidt-Hebbel, 1991).
Unit root tests conducted on the variables confirmed that most of the series have unit
roots of order one (i.e., I-I series). The small sample size for our estimation exercises
(22) suggest that a lot of intuition and judgement are called for in trying to draw
inferences from the cointegration and diagnostic tests. For example, the standard Dickey-
Fuller and Augmented Dickey-Fuller tests for order of integration and cointegration have
their critical values reported for sample sizes of at least 25 and 50, respectively. In spite
of this limitation, the procedure is still attractive because it imposes some discipline in
the estimation process. Following leads from the Masson, et al (1990), we proceeded
to estimate most of the equations in error-correction forms. It has been shown that as
long as such error-correction estimation is not strongly rejected by the time series
structure of the variables involved, it can improve the chances of estimating the true
parameters of interest. Most of the equations are in log-linear forms and the ordinary
least squares is used as the estimator. Most of the variables are expressed in real terms.
46 RESEARCI-1 PAPER 32
The estimation results for the stochastic equations are presented in Appendix 3. The
diagnostic statistics presented are the conventional ones: the t-statistics (in parenthesis),
the Schwarz Criterion (SC), R2, and the DW statistics (where relevant). Other diagnostic
tests conducted for each of the equations include a test for the presence of autocorrelation
and for the violation of homoscedasticity and normality assumptions. We also tested for
the stability of the estimated coefficients by applying the Chow tests. All the tests, taken
together, provide evidence that some of the individual equations may fairly characterize
the underlying data generating processes.
Evaluation of model simulation properties

The dynamic properties of the model are also evaluated. This is because, as Pindyck and
Rubinfeld (1981: 361) note, even if all the individual equations fit the data well and are
statistically significant, we have no guarantee that the model as a whole, when simulated,
will reproduce those same data series closely.
Methods of testing the overall quality of fit of general equilibrium models in any
rigorous way are still in their infancy ( Khan and Knight, 1991). Conventionally,
evaluations of macroeconomic models emphasize the tracking ability of the model, that
is, its ability to duplicate turning points. Statistics such as the mean-absolute-percentage
errors (MAPE) or the Theil's inequality coefficients are computed, and the ability of the
model's ex ante and ex post forecasts to track actual data closely is evidence of good
performance. As Hendry and Richard (1982:4) argue, however, selecting models by the
goodness of their dynamic simulation tracking performance is invalid in general and tells
more about the structure of the models than about their correspondence to reality. It
should be stressed that the choice of criteria for model evaluation depends crucially on
the purpose of the model (Pindyck and Rubinfeld, 1981: 361).
Some scholars believe that for models designed for policy analysis of the what if kind,
the model builder should ensure that: the individual equations are congruent models, and
the simulated system is dynamically stable and exhibits sensible simulation properties.
By this, it is required that the model should exhibit a tendency to return to equilibrium
after any shock, and the model's results after a shock should be qualitatively consistent
with theoretical priors. Since such model's are designed to ask what if questions, and
not specifically for forecasting purposes, some what if kind of shocks are applied to the
models and their results evaluated on the basis of their consistency with modelled
relationships.
Our model is designed for analysis of the "what if" kind of questions. However, in
spite of the seeming inadequacy of tracking performance as model evaluation criteria for
this kind of model, we are still sensitive to the size of the simulated model's residuals.
Tests on the actual and simulated baseline data for the period 1980-1990 show that the
simulated baseline data for some of the endogenous variables have a tendency to duplicate
MACROECONOMIC ADJUSTMENT, TRADE AND GROWTh 47
most of the turning points. Variables with the worst performance, at least as measured
by the root-mean squared percentage errors of their forecasts are mainly identities,
especially the debt variables. In particular, the stock of private external debt (PDEBT),
stock of official external debt (ODEBT), voluntary debt service payments (VOLSEV) and
real disposable income (YD) show root-mean squared percentage errors of 14.39, 12.84,
12.17 and 10.42, respectively. Surprisingly, the interest parity condition explained
exchange rate variable closely well, with a simulation percentage error of 7.82. All the
other variables have simulation percentage errors of less than 10. Furthermore, the
calculated eigensystem for the model shows that the model is dynamically stable, as the
eigenvalues for all the model's sub-sectors are less than unity in absolute terms.
Eigenvalues for the external debt sector and monetary sector are 0.93 and 0.91
respectively. Values for the other sectors lie between 0.79 and 0.88.
Policy simulations and analysis of
results
Specification of simulation rules
As indicated earlier, two versions of the model are created: one incorporates forward-
looking, model-consistent expectations, while the other assumes that economic agents
form their expectations adaptively. We then use two shocks that are critical for the
success of a structural adjustment programme in Nigeria to shed further light on the
properties of the model, as well as investigate the possible implications of these shocks
for macroeconomic stability and growth in Nigeria. The shocks are: a programme of a
permanent fiscal contraction, amounting to 2.5% of real GDP and an exogenous oil price
shock amounting to 10% permanent increase in the nominal dollar export- price of oil.
These shocks are designed to illuminate some of the issues that are of prime interest to
policy makers and also central to a successful adjustment programme6.
The two simulations are conducted for a period of 18 years, from 1993 through 2010.
They are as follows7:
1. A permanent and immediate cut in the real value of government expenditures

equivalent to 2.5% of real GDP. The cuts are assumed to be immediate in the sense
that they are not phased-in shocks, and the implementation begins in the same year
of the announcement, 1993. For economic agents in the model with forward-looking
expectations, the shock is unexpected and therefore comes to them as a surprise;
hence they have no time to act in anticipation of the fiscal contraction. Private
decision makers are assumed to regard the programme of deficit-reduction as fully
credible and after the initial surprise, begin to revise their expectations based on the
future cuts.
2. A 10% permanent increase in the exogenously determined export price of oil.

The shock is also assumed to be implemented immediately, starting in 1993.
It is important to make explicit assumptions about the stance of monetary policies

when conducting specific policy shocks since alternative assumptions will produce
significantly different results. In our simulations, monetary policy is assumed to be
unchanged in the sense that the target path for money stock is kept unchanged from the
baseline target path. As indicated earlier, the intertemporal closure rule assumes that the
income tax rate adjusts in order to prevent the stock of total government debt (domestic
and external debt) from rising without bound relative to GNP. This rule enforces, over
time, the long-run intertemporal budget constraints facing the government. As a policy
rule, this constraint requires that current budget practices of the government be
sustainable over time, and thus current debt liabilities must be serviced or repaid in the
future. The constraint thus estimates the adjustments in taxes required to sustain any
level of government consumption, given an exogenously specified target debt-to-GNP
ratio. Furthermore, the modelling of the government's fiscal behaviour effectively
enforces a fiscal and monetary rule that is analogous to operating an oil stabilization fund
with sterilization. This is because, in our model, increases in government revenues
resulting from increased oil exports or other sources are not spent; instead, the increased
revenues lead to increased servicing of existing stock of debt.
We have presented the simulation results in a standard format. Simulations using the
two model versions (that is, comparisons of the two expectations regimes) are conducted
only for the fiscal contraction shock. Results for the oil price shock are reported only
for the model-consistent version. This is because the qualitative differences between the
two model versions have been demonstrated with the fiscal shock, and we do not want
to repeat the comparison exercise for the oil price shock. For fiscal contraction shock,
detailed results are plotted in charts in a manner to assist quick comparisons of the results
of the two model versions. (The charts for the fiscal contraction simulations and the
results for the oil price shock are in Appendix 4). For the fiscal contraction shock, the
graph legends correspond to the two expectations regimes. For example, graph legend
ADAP always refers to the version of the model with adaptive expectations, while the
legend CONS always refers to the version with forward-looking, model-consistent
expectations. Simulation results reported for most variables are given as percent
deviations from baseline. Other variables are given as absolute deviations from baseline
(for example, trade balances, and interest rates).
Analysis of simulation results
Consequences of deficit-reduction shock

The results under the two expectations show marked similarity in their qualitative
outcomes but the magnitude of the multipliers and the timing of the effects show some
differences that cannot be ignored. While the long-term results are similar for the most
part, the short-run dynamic paths of some variables differ even though the policy shock
is unannounced. Of course, it is not relevant to simulate announced policy shocks with
adaptive expectations model because agents in the model lack the forward-looking
capability to react to future shocks until they are imposed. The driving force in the
observed differences between the two model versions is the differential behaviour of the
exchange rates. Economic agents in the forward-looking model expect a future exchange
rate appreciation to offset the persistent decrease in the short-term rate as a result of the
fiscal contraction8. As a consequence, exchange rate depreciates more under the model-
consistent version than the adaptive expectations.
As expected, the cut in government expenditure, by decreasing aggregate demand,
causes output, income and interest rates in both model versions to fall relative to baseline
in the short run. Output and income, however, fall less under the consistent expectations
because economic agents raise their expectations of future income. This induces private
consumption and investment to rise more under consistent expectations than adaptive
expectations. The reason is that consumers and investors project lower real interest rates
in the future, and thus higher future real wealth streams and investment profitability. In
the medium to longer term, income and output rise moderately above baseline. This is
caused by the crowding-in effects of domestic expenditures induced by lower interest
rates, prices and improvement in the economy's external competitiveness because of the
depreciation of the naira.
Interest rate variables also fall under the two expectations regimes, due to observed
contraction in the transactions demand for money as well as reduced pressure on available
money supply resulting from lower government borrowing in the domestic financial
market9. Another factor that could potentially depress the interest rate variables is the
behaviour of the risk premium term. As a result of the higher exports and lower stock
of government debt (domestic and external), the debt burden as well as the default risk
premium and the risk adjusted domestic interest rate should fall.
The initial fall in income and output depresses prices below baseline. Price inflation
is determined in an augmented Philips curve fashion, and movements of the economy
towards full employment accelerate movements in prices. The fall in prices relative to
baseline is modest, which is not surprising. This is perhaps caused by the offsetting
effects of the factors determining the price level. The fall in domestic income would
have depressing effects on prices whereas the depreciated naira, by raising import prices
would act to tug prices upwards.
Private investment falls relative to baseline in the short run, perhaps due to reductions
in government expenditures, which complement private investment spending, and the
high cost of imported inputs induced by the depreciated naira. The lower prices,
combined with lower interest rates, and increased private consumption spending drive
private investment above baseline after the first year under consistent expectations but
after two years under adaptive expectations.
The intertemporal fiscal closure rule enforced during the simulation results in the
drastic fall in both the income tax rate and income tax revenues. The fall in income taxes
combines with the fall in the price level to raise aggregate real disposable income sharply
and boosts private consumption expenditures10. Given the fall in long-term interest rates,
consumers generally increase their consumption in both versions of the model.
The export sector gains a boost. Oil export seems to gain the most. Oil exports are
expressed in real terms and given as an identity in both models. The change in oil
exports seems to be induced by changes in prices rather than any real change in quantity
exported. Since the shock does not induce changes in either the OPEC quota to Nigeria
or the dollar denominated export price of oil, the variables driving the increases in oil
export values are the depreciation of the exchange rate, which increases the nominal naira
value of oil exports and the fall in the price level. The real value of oil exports,
therefore increases even without any change in the exogenously determined export
quantity and price. Thus, as exchange rate approaches the baseline and the price level
rises above baseline in response to increases in output, real oil exports also fall to the
baseline. The exports of primary commodities and manufactures exports seem to follow
more the trends in domestic output than changes in exchange rate. These exports do not
rise significantly following the depreciation of the naira, which, by raising the naira value
of exports, should encourage the exporters. Supply of exports of manufactures and
primary commodities may therefore be constrained more by domestic supply capacity
than by price incentives. However, even though these export categories hardly moved
above baseline, it is encouraging to observe that a fiscal contraction programme in
Nigeria would have positive effects on exports generally.
Imports of goods and services declined relative to baseline in the short and medium
terms in both model versions. This is principally because of the decline in aggregate
demand and also the expenditure-switching effects of exchange rate depreciation. Imports
of consumer goods decline less than imports of capital goods and raw materials in both
models. Disposable income increases as a consequence of the reduction in the income
tax rate, and this induces more imports than would otherwise be the case. However, as
income and output recover from the initial fall and the exchange rate approaches baseline,
demand for all imports rises. In the long run, imports rise above baseline. Observed
trends in trade balance reflect the trends in imports and exports. Since exports are
dominated by changes in oil exports, changes in trade balance follow leads from oil
exports. Trade balance is positive throughout the simulation period, but improves more
under model-consistent expectations because of expenditure-switching effects of the
exchange rate depreciation.
The simulation results for the debt sector are interesting. The increase in exports and
output increases the government's debt servicing capacity. Given the debt service
reaction function, servicing of external debt improves relative to the baseline. The two
model versions assume that domestic debt is always fully serviced at given domestic short
term interest rate. Because the short-term rate falls relative to baseline, the total
contractual debt service payments (domestic and external) also falls. Given the reduction
in government fiscal deficits, and the improved debt service payments, the total stock of
debt (domestic and external) also falls relative to baseline. Improvements in
government's debt service behaviour lower the default risk premium on external debt,
restore external investors' confidence and lead to increased flow of financial resources
from private external lenders to domestic private sector.
Consequences of oil price shock

The results of a 10% increase in the exogenously determined price of oil are presented
in Figures 2a and 2b in the appendix. Simulation results have been reported for oniy the
model-consistent expectations version of the model.
Aside from fiscal discipline, another major issue in the discussions between the
government of Nigeria and the World Bank/IMF about maintaining macroeconomic
stability have focused intensively on how to manage shocks from oil earnings. The issue
is central to maintaining fiscal discipline since oil accounts for over 80% of government
revenues and 90% of foreign exchange earnings. What the government does or does not
do with the oil earnings is critical to the success of its overall economic management.
More importantly, since exogenous shocks in oil prices are mostly unpredictable, policy
discussions in recent times have focused on alternative scenarios for stabilization with or
without sterilization, so as to smoothen government expenditures along an exogenously
set target path and insulate the economy from the impacts (positive and negative) of the
oil shocks.
In the past, the Central Bank of Nigeria bore the responsibility of accommodating
fluctuations in foreign exchange flows. Oil related revenues (together with trade taxes,
etc.) were deposited in the Federation Account and shared among the three tiers of
government - federal, state, and local in a predetermined manner. Earnings from oil
exports therefore elicited corresponding naira deposits that were distributed among the
levels of government and some special funds. In the face of lax fiscal policy, and
without a sterilization programme, the policy of monetizing any windfall from oil
earnings increases money supply. Under a regime of flexible exchange rate and
competitive financial system, this would have the effects of lowering interest rates,
depreciating the exchange rate and causing inflation. The Central Bank's only recourse
to controlling money supply has been to contract credit to the private sector, since the
government maintains an unlimited overdraft with the Central Bank. Alternatively, the
Central Bank has resorted to panicky, direct control measures designed to mop-up excess
liquidity of the banking sector and causing the scarcity of loanable funds. This in turn
Sent interest rates to soaring as banks tried to make profits from limited lending, and the
spread between lending and deposit rates widened. Private sector activities therefore bore
the greater burden of adjustment to oil shocks, and got crowded out. So rather than
being a positive shock, most windfalls from oil were apparently managed in ways that
made them a 'curse'. The point being stressed is that the consequences of an oil price
shock depend on the policy regime in place.
If our focus were to recast history and capture the consequences of oil windfalls
during the last several years, then some modification of the present specifications have
to be undertaken. Greater disaggregation and even endogenization of components of
government expenditures would have to be undertaken. Money supply would have to be
fully endogenous in the model, and the oil earnings money supply nexus would need to
be spelt out. The other government's reaction functions (such as the central bank's credit
ceiling reaction function) would have to be modelled explicitly. Our framework in the
benchmark model does not aim to account for these linkages, since our emphasis is on
what would happen if government implements the new proposals under a new policy
environment? rather than with what happened in the past when the government
implemented certain policies?. No doubt, for some policy analysis, it would be beneficial
to endogenize components of government expenditure and experiment with specifications
of endogenous money supply mechanism. Our present effort does not cover those
specifications. Government expenditures (aside from the outlays for debt servicing) are
exogenous in the model, and money supply is treated as largely exogenous (even though
the central bank's interest rate reaction function is explicitly specified).
Our model and simulation exercise implicitly assumes a twin programme of
stabilization and sterilization. Stabilization is assumed in that the government devotes the
oil revenue windfall to the servicing of its debt and thus reduces the stock of debt.
Sterilization is assumed in that the simulation assumes that extra revenue resulting from
oil price increases does not lead to increased government expenditures nor increased
money supply. The exogenous component of government expenditure is not increased
as a result of the oil windfall. Our simulation results therefore demonstrate the potential
consequences of such policy directions for macroeconomic behaviour.
The essential channels through which changes in oil prices affect the domestic
economy are: impacts on government revenues, effects on total exports and thus on both
the voluntary external debt service payments and default risk premium on external debt
and changes in domestic income through impacts on trade balance.
Generally, the 10% increase in the nominal US dollar price of oil has beneficial
effects on the economy. A 10% increase in oil price (given an almost 100%
monetization into government revenues implied by our oil revenue equation) leads to a
10% proportionate increase in nominal oil revenues. But the depreciation of the
exchange rate increases the nominal naira value of oil revenues. As the government
revenues increase, the tax rate reaction function estimates a fall in income tax rates and
hence in the income tax revenues relative to baseline.
Increased oil exports as well as improvements in other export groups improve the
trade balance strongly and also positively affect domestic income. The increased exports
lower the debt burden, increase the government's debt servicing capacity and, given, an
unchanged monetary policy, lower the default risk premium on external debt. Risk
adjusted domestic interest rate therefore falls relative to baseline. The exchange rate also
depreciates. Exchange rate depreciation combines with increased aggregate demand to

push price levels above baseline.
Private consumption increases due to lower interest rates. Expectations of future
lower taxes raise the consumers' expectations of future real income and thus increase
their consumption. Private investment expenditures also increase in response to lower
interest rate and expected profitability of investment due to increased aggregate demand.
Manufactures and primary commodity exports hardly move above baseline because of the
very small changes in both aggregate demand and exchange rate.
Our model assumes that private agents in Nigeria face an endogenous supply schedule
for foreign loans that depends on the lenders' forward-looking assessment of the
government's debt servicing capacity. External debt owed by the domestic private sector
is assumed to be fully serviced. Foreign creditors, as risk averse investors, still look up
to Nigerian government's debt servicing records as a measure of the overall solvency of
the economy, and thus as the basis for extending loans to the domestic sector. An oil
price increase that increases total exports also increases government's debt service
performance. As a consequence, the flow of new external credit to the domestic private
sector increases substantially.
In conclusion, the simulation results show that a policy regime that enforces
stabilization with sterilization in the face of an oil export windfall would be generally
beneficial to the economy. The inflationary impact of the shock is minimal - hardly
above zero - but other macroeconomic aggregates show positive, albeit negligible,
improvements. To estimate fully the relative advantages or disadvantages of stabilization
and sterilization regimes as against the government's past fiscal and monetary responses
would require counterfactual analysis. In this case, one would be asking, what would
have happened if the government monetized the windfall, or spent it in some other
ways?. Results of such a hypothetical scenario would be compared with the results of
our regime that assumes sterilization of the windfall. It is intuitive to suggest that the
consequences of an oil windfall in the economy would depend on what the government
does with the windfall revenue.
IV Summary, conclusions and policy issues
In this study, a medium sized model of the Nigerian economy has been constructed and
focused to capture the essential elements of a structural adjustment environment. Two
versions of the model are created: one incorporates model-consistent, forward-looking
expectations in the assets and goods markets, while the other constrains private economic
agents to be adaptive in their expectations formation. Experiments with the two versions
are designed to highlight some of the subtle issues that differences in assumptions about
expectations formation might raise in evaluation of model results. Two illustrative
simulations are conducted to evaluate the model's simulation properties and to shed light
on the possible consequences of the much canvassed fiscal discipline, and also the
potential effects of an exogenous oil price shock.
Policy issues emanating from the simulation results
A major question that motivated the fiscal contraction shock is whether fiscal discipline
(a major policy proposal of the structural adjustment programme (SAP) ) is consistent
with the other broad objectives of the adjustment process, namely, macroeconomic
stability (especially price stability), promotion of exports and dampening of imports
through expenditure-switching effects of exchange rate changes and ensuring that the
Nigerian economy resumes growth in the medium to longer runs. In the last few years,
unsustainable fiscal budget deficits of the federal government have reached over 10% of
GDP mark, with destabilizing consequences in the economy. Some analysts argue that
reducing the fiscal deficit would depress the economy and be counter productive to the
objectives of SAP.
The model developed in this study is a highly simplified one, and is not adequate to
answer all the interesting questions that relate to a deficit reduction programme. For
example, no distinction is made in the model between different types of government
purchases, and government spending is treated as if it were all current consumption. All
investment expenditures are treated as private investment. There is no gainsaying the fact
that different government expenditures - be they capital investments or consumption -
have differential effects on short and long-run economic growth. Thus, depending on
which items are cut in the first instance, or whether there is a reallocation of expenditures
away from, say, consumption to investment spending, government fiscal behaviour could
produce markedly differential effects on the economy.
However, in spite of the highly aggregative nature of the model, important insights
can be gained from the experimental fiscal contraction shock. Some of the important
conclusions are:
1. A programme of budget-deficit reduction is consistent with the objectives of

maintaining price stability, improvements in external sector (depressing imports and
boosting exports), and ensuring medium-to long-term growth of the economy. It can
be observed that the short-run output and income loss (and perhaps the consequent
unemployment) may not be as big as some protagonists of the 'inevitability of
persistent deficits in order to maintain growth' would estimate. Indeed, if the budget
cuts are accompanied by accommodating monetary policy (a logical assumption), then
the estimated fall relative to baseline of income and output may be smaller, and the
growth impetus may be larger. Reduction in budget deficit would permit the Central
Bank of Nigeria to loosen the draconian squeeze on commercial bank credit to the
private sector and thus remove the crowding-out effects of government fiscal deficits
on the private sector.
2. The external trade sector improves moderately as a consequence of tightening the

fiscal laxity of the government. Exchange rate depreciation, combined with increased
domestic output, improves the external competitiveness of the export sector. Trade
balance is positive throughout the simulation period, and this contributes to the
positive growth of income.
3. The importance of price stability for economic growth, and for promoting other
macroeconomic objectives of the government, cannot be overemphasized.
Macroeconomic stability ensures that the private sector is protected from uncertainty
and shocks, and this allows banks to develop a healthier portfolio. From a policy
perspective, stability of the macroeconomic environment is a sine qua non for the
sustainability of financial sector liberalization. Combined with lower and stable price
inflation, competition among financial institutions would lower both the deposit and
lending rates. Depositors would demand lower risk premiums than they would
otherwise. Competition in the credit market would then promote the efficiency of
investment and thus growth. Simulation results obtained for the fiscal contraction
shock is consistent with these observations. Lower prices and nominal interest rates
result in lower long-term interest rates, and thus a boost of private investment
expenditure.
4. Differences between the short-run costs and long-run benefits of the fiscal deficit
reduction point to the trade-off involved in deciding on a programme of deficit
reduction. Different deficit reduction plans are conceivable, and each of them would
involve different 'mix' of costs and benefits'1. For example, a full drawn-out,
phased-in expenditure reduction spanning several years such as the one proposed by
the transitional government of Nigeria can be undertaken. However, if a major
segment of the private decision makers are forward-looking and regard the policy
announcement as fully credible, the negative consequences of such phased-in deficit
reduction may be less than expected. On the other hand, the government may decide
on an accelerated deficit reduction within a short period of time. Depending on the
composition of government expenditures and the items affected most by the cut, the
negative impacts could be more or less than expected. Whatever the nature and mix
of deficit reduction plans that are deemed most desirable, the critical element is the
political will to undertake such tough measures. Short-run depression, which could
accompany such a policy, may be so destabilizing politically that most governments
(especially ones with a fragile political base) would be reluctant to undertake such
policies. If the military regime in Nigeria could not muster the courage to discipline
government spending, it is doubtful that a civilian administration (which is not
generously endowed with such political courage) could conceivably embark on it, or
do so in a manner to warrant the desired long-run benefits.
The results of the oil price shock are revealing in a number of other ways. The
results show that an oil price shock (an increase in prices), aside from being inflationary,
can have beneficial effects on a broad range of economic aggregates. But these effects
are realized under an assumption that monetary policy and government spending are not
changed from their baseline paths as a result of the shock. In other words, by assuming
no changes in government's exogenous expenditures and money supply, our model
effectively imposes a kind of sterilization regime. However, when government spends
the windfall, and the Central Bank attempts to control money supply by crowding-out the
private sector, the consequences may not be as encouraging as our simulation results
show. This raises the question of how to cushion government spending in the face of
unpredictable, and sometimes frequent, oil price shocks.
The federal government of Nigeria seems to be aware of the necessity for some sort
of a stabilization mechanism and since 1989 has tried to implement a stabilization fund
for oil revenues. The programme is intended to achieve two major objectives: first, to
stabilize economic activity and smooth fiscal spending by increasing reserves (or
decreasing debt) in times of high oil revenues and by running down reserves when oil
revenues decline; second, to sterilize money supply increases resulting from the increased
oil revenues by not spending the naira counterpart funds of new reserves (stabilization
with sterilization). Our simulation results imply that both regimes- stabilization with
sterilization are imposed, in the sense that the windfall does not lead to increases in
government expenditures and money supply, but rather results in increased debt service
payments and thus reduces the stock of government debt.
One major obstacle to a credible implementation of a stabilization scheme is the lack

of political will to resist spending. For example, in 1990 the federal government
reserved much of the oil windfall resulting from the Iraqi war in its stabilization account.
In late 1990 the government spent much of the naira counterpart fund, and this increased
domestic money supply by an unprecedented 40% during the year in the face of
unchanged reserves. The naira depreciated rapidly in 1991 and inflation soared.
Reneging on announced or agreed policies is a major failure of the government of
Nigeria in its attempt to structurally adjust the economy. Our simulations show that a
stabilization and sterilization programme in the face of an oil windfall could be potentially
beneficial to the economy. Failure to fully implement such a policy regime therefore is
not a good testimonial for the government's economic management skills.
Expectations formation and policy choices

Simulation results reported for the fiscal shock in this study show that the manner of
expectations formation by private economic agents can affect the nature of the results
obtained in any simulation exercise. For example, when agents are forward-looking in
their expectations, a deficit reduction shock would dampen output for only two years
whereas the depression of output is more prolonged if agents are backward-looking. The
qualitative results (i.e., the overall direction of impacts) are comparable, but the
differences in the size of multipliers are too significant to be ignored.
The development of rational expectations assumptions in macroeconometric modelling
has come a long way. Protagonists of rational expectations have vigorously criticised the
clearly inadequate treatment of behaviour as being solely backward-looking, and
demonstrated the superiority of forward-looking behaviour as an alternative reality. It
is assumed that movements of the supply, demand and price relationships depend on
expectations of the future. For example, if investment and consumption depend on real
interest rates, then they implicitly depend on the expected price level in the future. Also,
long-term interest rates depend on expectations of future short-term interest rates, and
therefore, on expectations of future monetary and fiscal policies. If interest rate
differentials depend on expected changes in exchange rates, then expectations of future
exchange rates affect interest rate behaviour. Therefore, private economic agents who
are forward-looking in their expectations would take into account the likely effects of
future policies in their present behaviour (Soludo, 1992).
In general, however, the effects of policy shocks when agents are forward-looking
depend on the size of the shock, whether the policy changes are announced or
unannounced and whether the shock is persistent or temporary. When policy changes are
perceived as temporary, rational economic agents regard them as transitory and may not
react to them as much as they would if the shock is sustained. This is one reason the
reaction formation of economic agents under model-consistent expectations would dampen

the effects of temporary shocks and accelerate the effects of sustained policy changes.
What is the implication of these alternative treatments of expectations for model
selection and interpretation of model results for policy makers and analysts in a
developing country such as Nigeria? Put differently, given the rather heroic assumptions
underlying model-consistent expectations, are the assumptions relevant for policy analysis
in Nigeria? Definitive answers to these questions cannot be given. On one hand, it is
easy to dismiss rational expectations (which presupposes credibility of announced
policies, and perfect information) as inapplicable in Nigerian circumstances of
uncertainties, structural transformations and mistrust of government policies. On the other
hand, to simply assume (as most people are used to doing, sometimes even without a
thought about the implications of such assumptions) that the economic agents in Nigeria
are so naive and so stuck with their past that they hardly take account of the future in
their present economic calculations, is fantastic. Backward-looking behaviour implies
that the past is the best predictor of the present and the future. In circumstances of
structural transformation, irregularities and uncertainties about policy directions, reliance
solely on the past as a guide to economic behaviour is not realistic.
It may also not be realistic to ignore the anticipatory behaviour of many economic
agents especially in the financial, distributive and productive sectors of the economy.
Some firms are known to value their assets at their expected replacement cost rather than
the actual (past) cost. Pricing behaviour of some firms (especially ones that depend
heavily on imported inputs) can be realistically characterized as depending on the
expected future cost of inputs, and thus on some expected future exchange rate. When
firms decide to either accumulate or decrease stocks in anticipation of future policy
announcements in the annual budgets, they cannot be said to be solely backward-looking.
In Nigeria, past experiences show that prices have jumped in anticipation of policy
changes, especially when government announces wage increases. One may argue that
the size of economic agents exhibiting these anticipatory behaviours is small relative to
the economy. However, the impacts of their behaviour may not he too trivial to be
ignored by policy makers and analysts.
These expectations regimes are nevertheless extremes, and the actual behaviour is
probably somewhere in the middle. It is intuitive that economic agents take their
decisions on the basis of both hindsight and foresight. It is suggested that future
advances in macroeconomic modelling incorporate ideas about appropriate strategies to
construct the half-way house between these extreme expectations regimes.
In the present study, experimentation with the two expectations regimes is more for
demonstrative purposes than an attempt to prove that one extreme assumption is superior
to the other. If the exercise serves any purpose at all, it is to sensitize policy makers and
analysts about the implications of entertaining either of these assumptions in formal
models, and therefore to be cautious in interpreting multipliers generated by either of
these model versions. For practical purposes, since these expectations are extremes and
thus bracketing the reality, policy makers can realistically infer that the true behaviour
is somewhere in between the two results. A third purpose is to flag the issue for
macroeconomic analysts, especially those who construct and use economy-wide models,
and hope that further research efforts can be devoted to improving economists'
characterizations of the economy through models.
Aside from the policy issues illuminated by the representative simulations reported
in this study, it is important to note that the major contribution of the work lies more
with the potential uses of the model than what it has been used to do in this study. As
indicated earlier, the model attempts to capture the essentials of the SAP environment in
Nigeria, and thus some of the current issues in the adjustment process can be analyzed
with the model.
Potential uses of the model

The model is to be used as tool-kit, and can be potentially useful to the staff of the
Central Bank of Nigeria, Ministry of Finance, Budget and Planning Ministry, and policy
analysts. With some skills, the model can be extended and simulated to illuminate some
issues of concern to policy makers. The model-consistent version, for example, can be
used to investigate possible implications of announced versus unannounced policy shocks.
As Nigeria makes a transition to civilian rule, the importance of announced versus
unannounced policies cannot be ignored. This is because announced policies would
usually take time to be approved by the National Assembly, and it is not unreasonable
to expect economic agents to act in anticipation of the effects of the policies. Phased-in
policies, such as the budget deficit reduction plans announced by the transitional
administration, can be analysed with the model. A major debate has lingered on about
the potential consequences of removing the subsidy on petroleum products in Nigeria.
Whereas the model is not designed to handle distributional problems arising from such
a policy, it is adequate to illuminate some of the short-and long-term macroeconomic
effects of such a policy shock.
One major strength of the model is that it is flexible, and several policy rules can be
entertained as simulation rules. Intertemporal sustainability of policy regimes is a crucial
feature of the model. When the intertemporal closure rules are enforced, the model is
flexible to estimate (for use of policy makers) the necessary adjustments (costs) of
alternative policy scenarios. That is, the model shows what has to adjust in order for a
particular policy action to be sustainable. The treatment of intertemporal closure rules
is one of the novel aspects of the model, and it is a special feature of model-consistent
models.
Limitations of the model and agenda for further

research
It is unrealistic to aspire to build a 'super-model' of a developing economy - one that

captures all reality and leaves no room for further modifications. As every economist
knows, given the present state of knowledge in the economics profession, every effort
in building a general-equilibrium model of an economy is, like every work of human
creation, an unfinished business. Every effort should be seen as a stage in a process, but
the model-builder must be satisfied that each stage fulfils definite purposes, and deepens
understanding of some aspects of the economy in question. These caveats have
influenced our current efforts.
Every model reflects the builders' perceptions of the economy being modelled.
Economists are not agreed about the correct characterization of the workings of
economies, and this is also reflected in the assumptions we make in the process of
economic analysis and model-building. Our methodological inclination in building the
model is to start with a small model whose results we can understand. In the process,
we have made several assumptions about economic behaviour in Nigeria that contradicts
historical experience but are nonetheless analytically helpful and revealing.
It should be emphasized that revisions and extensions of the model will be a continuous
process. As it is now, the model is in a highly aggregative form and has several
deficiencies. Even though we tried to graft both the neo-classical and Keynesian
frameworks into the modelling process, the Keynesian bent of the model is evident.
Furthermore, we recognize the need for further research and refinements of the supply
sector; greater disaggregation and alternative modelling assumptions of the monetary
sector; better treatment of the external sector (imports and exports); and alternative
modelling of exchange rate behaviour. The external debt sector of the model can be
refined to take account of issues related to debt buy-backs, valuations in debt prices in
the secondary debt market, etc. Experiments with disaggregation and endogenization of
different government expenditure components can also be undertaken. Some of these
improvements are taken up in a follow-up study sponsored by the AERC and entitled
"Choice of Optimal Mix of Fiscal and Monetary Policy Rules: Evidence from A Model
of Nigeria".
Notes
See the 'Introduction' in Khan, Montiel' and Haque (1991) for detailed discussion
of the elements of SAP.
2. We are aware of the current efforts by the CEAR group to develop a model as input
into the Perspective Planning efforts of the Nigerian federal government, and a
model of Nigeria also being developed by the World Bank. However, we are
informed during preliminary discussions with some of the scholars involved in these
projects that the models emphasize the real sector and the financial sector is largely
ignored. This bent reflects the specific objectives of the modellers.
3. This is the actual debt service payments made by the government to service its stock
of external debt. It is not necessarily equal to the contractual obligations because the
government may not meet the contractual payments, and the difference is the interest
arrears, which is often capitalized. The external debt section of this paper provides
further duscussion on this variable.
4. This is similar to the specification by Meredith (1989-20). However, instead of a

bond maturity period of ten years, a period of eight years is assumed in this model.
This approximates the maturity period of the development stocks in Nigeria.
5. This reaction function is similar to the one specified for the EMS countries in
Masson, et al (1990), which is also a variant of the rule used by Edison, Miller and
Williamson (1987). Investigating the implications of a version of the model with the
assumption of a target-zone fixed exchange rate regime and the corresponding
monetary policy reaction function is a potentially beneficial line of research. The
current model does not investigate those, hut instead assumes a flexible exchange rate
regime.
6. For example, the fiscal indiscipline of the Babaginda administration is widely

acknowledged as the bane of the current adjustment sprocess in Nigeria. The
unsustainably burgeoning fiscal deficit is seen as a major destabilizing factor in the
economy and has been a source of concern to both domestic economic agents and
external investors and donor agencies. It is therefore not surprising that the
transitional governemtn headed by Mr. Shonekan announced a long-term programme

to reduce the fiscal budget deficit to about 3% of GDP by 1997. Since the present
study is designed to evaluate the consistency of some of the proposals for fiscal
discipline under SAP, with other objectives of promoting trade and growth, a budget
deficit reduction shock is considered a reasonable experiment.
7. The simulations are conducted with PCSIM, a Fortran-based software currently being
maintained at the Brookings Institution, Washington, D.C. It should be observed
that solving large non-linear models incorporating forward-looking, model-consistent
expectations can entail technical complications. The procedure used to solve the
model is an application of the Fair-Taylor extended-path technique (see R.C. Fair
and J.B. Taylor, 'Solution and Maximum Likelihood Estimation of Dynamic
Nonlinear Rational Expectation Models', (Econometrica 1983).
8. Recall that interest parity condition adjusted for default premium on debt stock
determines exchange rate in the model, and an absence of capital controls is
assumed. Therefore, the vairable that adjusts to ensure zero balance in the balance
of payments is primarily private capital flows.
9. This result is intuitive given the assumptions underlying the modelling of the
monetary sector of the model. The Central Bank is assumed not to purchase
government bonds, and so fiscal deficits must be financed by borrowing from the
domestic private financial institutions and abroad. Deficit financing therefore raises
domestic interest rates and deficit reduction is expected to have opposite effects in
the model, given a simulation rule of 'unchanged' monetary policy. This is a
realistic rule to enforce given the current prosposals for deficit reduction and indirect
monetary policy instruments in Nigeria.
10. It should be observed that different specifications of this intertemporal closure rule
would produce results that may be significantly different from some of the results
obtained from the simulation - at least the size of the multipliers may be significantly
affected.
11. We plan to experiment with alternative simulations of the different deficit-reduction

programmes in the future. This will be done with versions of the model that
incorporate certain revisions and improvements on the model structure and
properties.
References
Adam, C.S. 1992. "Lecture Notes for Technical Workshop on Time Series
Econometrics", Nairobi, 23 September - 3 October 1992.
Adams, C and Adams C .11. 1990. "A Scenario and Forecast Adjustment Model for
Developing Countries", in World Development Outlook. Washington, D.C:
International Monetary Fund.
Adamson, Y.K. 1974. A Linear Short-term Forecasting Model for National Economic
Planning in Mgeria. MSc dissertation, Birmingham University, U.K.
Ahmed, S. and Chhibber, A. 1989. "Non-Inflationary Growth With Debt Management:

Indonesia's Achievements and Options", mimeo, The World Bank.
Allen, C. and Vines, D. 1992. "Some Key Global Adjustment Scenarios and Their
Effects on Major Developing Country Regions", A Report to the United Nations,
June.
Aziz, U.A. ed. 1990. Strategies for Structural Adjustment: The Experience of Southeast
Asia. Washington, D.C : IMF/Bank Negara Malaysia.
Bacha, E, and Edwards S. 1988. "Seminar on Macroeconomic Adjustment and Growth",

Journal of Development Economics, vol. 29.
Beenstock, M. 1991. "An Econometric Model of the Oil Importing Developing

Countries", Paper presented at the Conference on Global Interdependence: Theories
and Models of the Linkages Between OECD and Non-OECD Economies, held in
Seoul, South Korea, May 27-28.
Blinder, A.S. and Deaton. A.C. 1985. "Time Series Consumption Function Revisited",
Brookings Papers on Economic Activity, 2, pp. 465-511.
Boltho, A. 1988. "Can We Return to Rapid Growth?", World Bank PPR Working Paper,
9, Washington, D.C.
Bryant, R.C. 1992. "Comment" on the paper, "Project Link: Past, Present and Future",
by B.G. Hickman and K.G. Ruffing.
Camdessus, M. 1987. "Opening Remarks" in V. Corbo, M. Goldstein, M. Khan, eds.

Growth-Oriented Adjustment Programs., Washington, D.C.: International Monetary
Fund/The World Bank.
Central Bank of Nigeria. 1988, 1990, 1992. Annual Report and Statement of Accounts,
Lagos.
Chadha, B. Masson, P. and Meredith, G. 1991. "Models of Inflation and the Costs of
Disinflation", JMF Working Paper, International Monetary Fund.
Chhibber, A. and Shafik, N. 1990. "Does Devaluation Hurt Private Investment?: The
Indonesian Case", mimeo, The World Bank.
Dittus, P. O'Brien, P.S. and Blommestein, H.J. 1991. "International Economic Linkages
and the International Debt Situation", OECD Economic Studies, No. 16 (Spring).
Edison, H.J. Miller, M.H. and Williamson, J. 1987. "On Evaluating and Extending the
Target Zone Proposal", Journal of Policy Modeling, vol. 9 (Spring), pp. 199-224.
Ekeoku, N.I. 1984. "Simulation and Fiscal Policy Analysis with a Macroeconometric
Model of the Nigerian Economy", PhD thesis, University of Ibadan.
Elbadawi, A.I. and Schmidt-Hebbel, K. 1991. "Macroeconomic Structure and Policy in

Zimbabwe: Analysis and Empirical Model (1965-88)", Policy, Research and External
Affairs Working Papers, WPS 771, The World Bank.
Engel, R. and Granger C. 1987. "Cointegration and Error-Correction: Representation,

Estimation and Testing", Econometrica, 55, pp. 251-276.
Fair, R.C. and Taylor, J.B. 1983. "Solution and Maximum Likelihood Estimation of
Dynamic Nonlinear Rational Expectation Models", Econometrica.
Federal Republic of Nigeria. 1990. First National Rolling Plan 1990-1992, vol. 1,
Federal Printing Press, Lagos.
Friedman, B.M. 1991. "How Does It Matter", NBER Working Paper No. 3929
(December). Cambridge: National Bureau of Economic Research.
Goldstein, M. 1986. "Global Effects of Fund-Supported Adjustment Programs", IMF

Occasional Paper No. 42, Washington D.C.: International Monetary Fund.
Goldstein, M. and Khan, M.S. 1982. "Effects of Slowdown in Industrial Countries on

Growth in Non-Oil Developing Countries", IMF Occasional Paper No. 12,
Washington D.C.: International Monetary Fund.
Granger, C. and Newbold, P. 1974. "Spurious Regressions in Econometrics", Journal

of Econometrics, 2, pp. 111-120.
Harris, J.R. 1985. "A Survey of Macroeconomic Modelling in Africa". Paper presented
at a meeting of Eastern and Southern Africa Macroeconomic Research Network,
December 7-13, Nairobi, Kenya.
Hendry, D. and Richard, J. 1982. "On the Formulation of Empirical Models in Dynamic
Econometrics", Journal of Econometrics, 20, pp. 3-33, North-Holland Publishing
Company.
Khan, M. and Knight, M. 1985. "Fund-Supported Adjustment Programs and Economic

Growth', IMF Occasional Paper No. 41, Washington, D.C.: International Monetary
Fund.
Khan, M. and Montiel, P. 1989. "Growth-Oriented Adjustment Programs: A Conceptual

Framework", Staff Papers, June, Washington D.C.: International Monetary Fund.
Khan, M., Montiel, P. and Haque, N. eds. 1991. Macroeconomic Models for Adjustment
in Developing Countries. Washington D.C.: International Monetary Fund.
Lahiri, A.K. 1989. "Dynamics of Asian Savings: The Role of Growth and Age
Structure", Staff Papers, Washington D.C.: International Monetary Fund, vol. 36, pp.
228-261.
Lipumba, N., Ndulu, B., Horton, S. and Plourde, A. 1988. "A Supply Constrained
Macroeconometric model of Tanzania", Economic Modeling, pp. 354-375.
Masson, P., Symansky, S. and Meredith, G. 1990. "MULTIMOD Mark II: A Revised
and Extended Model", IMF Occasional Paper No.71, Washington, D.C.:
International Monetary Fund.
McKibbin, W.J. 1991. "A Multi-Sector Growth Model of the World Economy Focusing
on the Linkages between the OECD and Non-OECD Economies", Seminar paper
presented at the World Bank, October.
Muscatelli, A. and Vines, D. 1989. "Macroeconomic Interactions Between the North and
South", in R.C Bryant, D.A. Currie, et al (eds.) Macroeconomic Policies in an
Interdependent World", Washington, D.C.: The Brookings Institution.
Nelson, R. and Plosser, 1982. "Trends and Random Walks in Macroeconomic Time
Series: Some Evidence and Implications", Journal of Monetary Econometrics, 10, pp.
139-162.
NISER. 1983. The NISER Econometric Model of the Nigerian Economy, Nigerian
Institute of Social and Economic Research, Ibadan.
Ojo, G.O. 1972. "An Econometric-Simulation Model of the Nigerian Economy", PhD
thesis, State University of Science and Technology, Ames, Iowa.
Olofin, S. 1985. Modelling Nigeria's Economic Development, Ibadan University Press,

Ibadan.
Olofin, S. 1977. "A Tentative Specification of an Econometric Model of the Nigerian

Economy", mimeo, University of Ibadan.
Onitiri, H.M.A. 1992. "Devaluation and Structural Adjustment: The Nigerian

Experience", Development Research Paper Series, No. 3, United Nations Economic
Commission for Africa, Addis Ababa.
Pindyck, R. S. and Rubinfeld D . L. 1981. Econometric Models and Economic Forecasts.

London: McGraw-Hill Book Company.
Poloamina, I.D. 1986. "An Experimental Ex-post Analysis of the Effectiveness of Fiscal
and Monetary Policy Measures in Nigeria's Development Plans (1975 -1985)" PhD
thesis, University of Ibadan.
Research Department, IMF. 1987. "Theoretical Aspects of the Design of Fund-Supported

Adjustment Programs", Occasional Paper, No. 55, Washington, D.C.: International
Monetary Fund.
Sachs, J.D. 1987. "Trade and Exchange Rate Policies in Growth-Oriented Adjustment
Programs" in V. Corbo, et al, eds. Growth-Oriented Adjustment Programs.
Washington, D.C.: International Monetary Fund/The World Bank.
Salim, A. S. 1990. Africa: The challenges of European Economic Integration. The 1990
Guardian Lecture. Lagos: The Guardian Publishers.
Salvatore, D., ed. 1989. African Development Prospects: A Policy Modeling Approach.
New York: Taylor and Francis.
Soludo, C.C. 1993. "Simulating Global Policy Scenarios: The Effects of Alternative LDC
Closure Rules", Paper presented at a CEPR international conference on
Macroeconomic Interactions Between North and South, held at the Institute of
Statistics and Economics, University of Oxford, U.K, April 23-24.
Soludo, C. C. 1992. "North-South Macroeconomic Interactions: Comparative Analysis

Using the MULTIMOD and INTERMOD Global Models", Brookings Discussion
Papers in International Economics, No. 95.
Sundberg, M. 1988. "The Transmission of Fiscal Policy Shocks to Developing Countries:

Preliminary Simulation Results", mimeo, June.
Sundberg, M. 1990. "The International Transmission of Macroeconomic Policy

Disturbance Among Countries of the Pacific Basin", Draft PhD thesis, Harvard
University.
Svensson, L.E.O. 1992. "An Interpretation of Recent Research on Exchange Rate Target
Zones", Journal of Economic Perspectives, vol. 6, Number 4, pp. 119-144.
Symansky, S. 1991. "Closure Rules and A New Developing Country Region for
MULTIMOD". Paper presented at the conference on Global Interdependence:
Theories and Models of the Linkages Between OECD and Non-OECD Economies,
held inSeoul, South Korea, May 27-28.
UNCTAD. 1973. "Model of the Nigerian Economy", Ball (ed.) The International
Linkage of National Economic Models. Amsterdam: North Holland Pub. Coy.
Uwujareen, G.P. 1977. "Specification and Estimation of an Economy-wide

Macroeconometric Model of Nigeria", The Nigerian Journal of Economic and Social
Studies, vol. 14, No. 3.
World Bank. 1974. "World Bank Long-term Projection Model of the Economy", in
Nigeria, Options for Long-term Development. Baltimore: Johns Hopkins Press.
Appendix 1. Data sources and limitations
Care has been taken in the collection of data to use specific sources for different kinds
of data in order to ensure consistency. As much as possible, we have relied on
publications of the Central Bank of Nigeria for data on the financial sector, interest rate,
exchange rate, domestic debt, output, government finances and external trade. Such
publications include the Annual Report and Statement of Accounts, Nigeria's Principal
Economic and Financial Indicators, and particularly, the Statistical Bulletin (volumes 1
and 2, 1990 and 1991). The World Bank's World Tables, World Debt Tables and the
IMF's International Financial Statistics are the major sources of data on external debt.
Data on oil production, export, domestic consumption and export prices are sourced
from the Annual Statistical Bulletin of the Nigerian National Petroleum Corporation.
Historical nominal prices of some petroleum products were obtained from the sales files
of some sampled petrol filling stations. Other supplementary sources of data are from
publications of the Nigeria's Federal Office of Statistics: Annual Abstract of Statistics;
Digest of Statistics; and Economic and Social Statistics Bulletin. Labour statistics, among
others, are sourced from these publications.
It can reasonably be argued that the quality of a model's simulation results is as good
as the quality of data used in calibrating the model. This can be particularly true for
developing countries where the quality and availability of macroeconomic data are highly
suspect. In Nigeria in particular, disparities in the values of data collected from different
sources, and the absence of data series for several years, make macroeconomic research
a daunting task. For many series, blank columns for several years are a common feature
of major statistical publications. Collecting macroeconomic data often entails personal
visits to the agencies responsible for documenting information on these series, and it can
be a lot of effort trying to extract the needed data from files, reconcile the numerous
inconsistencies and extrapolate some series to fill blank columns.
Furthermore, the incomparability of data from different sources, or even data from
the same source but published at different periods, is a frustrating experience. A lot of
care and discretion are therefore needed in deciding which sources to rely on for different
series. In spite of all the caution exercised, it is still difficult to vouch for the quality or
reliability of the data series used in the study. Multipliers emanating from our
simulations, and indeed from most macroeconomic analysis with Nigerian data, have to
be taken with caution.
Macroeconomic research cannot, however, be halted in many countries because of bad
data. Policymaking and policy analysis still go on in these countries on the basis of the
same bad data. There is too much at stake for economists to simply wring their hands
in despair because of bad data. Indeed, macroeconomists in developing countries can

play a critical role in ensuring the evolution, in the longer-term, of better quality data
bases by interacting actively with the data collection agencies, pointing out which data
are needed and in what form. Basic errors and inconsistencies in data series should be
continually brought to the attention of the agencies, and this positive interaction may
result in greater efficiency over time. In the long run, quality of data may be improved.
Improvement is the goal since quality of macroeconomic data in general is a matter of
degree.
Aside from quality of data, an important decision has to be made with respect to
methodology for extrapolating data beyond the period for which they are available. This
is an important stage, but one that calls for a lot of judgement. The only guide to
preparation of out-of-sample data bases is one's judgement about future course of
policies, developments in the economy or assumptions about what 'sensible' future paths
of economic behaviour would look like.
A number of decisions have been made about post-sample projection of data. First,
we chose to limit the projection of data into the future to the year 2015. Even though
the period is not long enough for investigations of the model's steady state properties, we
believe it is enough to make inferences about the model's long-run simulation behaviour.
Second, a decision was made to distinguish between periods in the future data series.
There is a transitional period that spans 1993 through the year 2000, and a second phase
that runs from 2001 through 2015. During the transitional period, a three-year moving
average growth rate of GDP, imports, exports, consumption and private expenditures is
assumed to continue. From the year 2001, the economy is projected to grow at an
annual rate of 3.5%, and other major macroeaggregates are assumed to maintain their
constant ratios to GDP. Deficit reduction programme is assumed to be enforced in the
remainder of the 1990s, with the government running surpluses in the several years after
the year 2000 and then balancing its budget annually. In the longer term, a zero trade
balance is assumed after a sustained period of trade surpluses to offset earlier deficits.
The exchange rate is projected to stabilize around 25 naira to a US dollar. Prices and
interest rates are projected to stabilize after the year 2005, and stable but positive real
rates to be maintained thereafter. All these assumptions, though arbitrary, are predicated
on the need to satisfy intertemporal behavioural constraints. The government cannot
borrow forever, and accumulated debts have to be serviced. Budget or trade deficits
today have to be offset by surpluses tomorrow. The choice of growth rates may be
arbitrary, but once chosen, other variables are projected in a manner consistent with the
assumed growth rates of certain aggregates.
Appendix 2. List of variables
Endogenous variables
AUR Average interest rate on the two sources of external debt

CAIMP Capital goods imports
CAPUTL Capacity utilization rate
CNIMP Consumer goods imports
CP Private consumption expenditures
COMTAX Company or corporation income tax
CURBAL Current account balances (in naira)
CUSTAX Customs and excise duties
DARREA Accumulated arrears on external debt service obligations
DDEBT Stock of government's domestic debt
DPIT Total imports price in domestic currency.
DPXT Total exports price
DRP Default risk premium on external debt
DSERV Service payments on domestic debt
EDEBT Total stock of external debt
ER Nominal exchange rate (naira per US dollar)
Expected nominal exchange rate
EXP Total real exports
GDEF Government nominal current budget deficit
GDP Real gross domestic product
GEXP Total nominal government expenditures
GNP Real gross national product
GTAX Total nominal government revenues
GTAXF Total federally retained revenues
IMP Total real imports
INF Inflation rate
INFe Expected inflation rate
INV Private investment expenditures
KP Gross private capital stock
LF Demand for labour force
LR Long-term (lending) interest rate
M Demand for real money balances
MAEXP Export of manufactures and semi-manufactures

NFA Net foreign assets (in naira)
NFPCAP Private to private capital flows
NPDEBT New inflow of external private lending to the government
ODEBT Stock of external debt owed to official sources
ODTSEV Contractual debt service obligations on official debt
OILEXP Total oil export receipts
OILTAX Revenues from oil sources (petroleum profit tax)
OTHOIL Other oil related revenues
PDEBT Stock of external debt owed to private creditors
PDTSEV Contractual service obligations on debt owed to external private
creditors
PGNP Domestic price level (GNP deflator)
PMEXP Export of primary commodities
REXR Real exchange rate
RLENR Real long-term interest rate
RSR Real short-term interest rate
DRS Short-term interest rate
TASETS Total external assets (private and public)
TBAL Trade balance (in local currency)
TDEBT Total government debt stock (domestic and external debt)
TDTSEV Total contractual debt service obligations on external debt
TRATE Average income tax rate
TSERV Total debt service payments (domestic and external)
UE Unemployment rate
VOLSEV Voluntary or actual debt service payments
WG Average wage rate
YCAP Capacity (potential) output
YD Real disposable income
YTAX Nominal income taxes
Exogenous variables and parameters

Dep. Depreciation rate on capital stock
Dom. Proportion of government current deficits financed by borrowing in the
domestic market
DPIM Exogenously determined price of imports (valued in dollars)
DPOIL Price of oil products in the domestic market
DUM Dummy variable used to adjust tax rate reaction function
Fed. Share of the federal government in the federation account

FUR Fixed interest rate on external debt owed to official sources
G Exogenous real government expenditures
GFA External assets of the government (in naira)
LF Supply of labour force
MT Target money stock (money supply)
NODEBT New inflow of external loans to government from official creditors
PCOM Primary commodity prices
PCR Banking sector credit to the private sector
PFA Private sector's stock of foreign assets
POlL Export price of oil (in US dollars)
PQOTA OPEC determined oil production quota for Nigeria
T Trend term that captures the effect of technical progress
TarifM Average tariff rate on imports
TarifX Average tariff rate on exports
TDEBTT* Exogenous target level of government debt (domestic and external)
TRASF Government transfers
URS US short-term interest rate, also used for calculating debt interest
payments
URSR US real short-term interest rate
Appendix 3: Estimation results
Results of the estimated equations are presented below. Figures in parenthesis are the
t-statistics, a is the equation standard error, SC is the Schwarz Criterion, and R2 is the
usual R squared statistics.
1. Private Consumption Expenditure:
CP YD (YDtl)
Ln ( -0.0117 +0.639 Lii ( )+0.087 Liz —0.2677 RLR
cPt_1 )vt_.I cPt_I
(0.78) (5.74) (2.08) (1.95)
R2 = 0.88 a = 0.073 SC = -4.76
2. Private investment expenditures:
Ln(INV) = 2.58 + -0.0936Ln(REXR) -

(1.88) (3.31) (1,79)
-0. 1027Ln(RLENR) + 0.
(1.53) (3.91) (1.97)
R2 = 0.96; DW= 2.28.

3. Oil tax revenue:
OIL TAX OILEXP * PGNP

Liz ( ) - 0.027÷0.937 Ln ( +
OILTAX : -1 OILEXP : 1PGNP : -i
(1.12) (14.93)
PGNP
0.084 Ln (OILEXP*
OILTAX
(1.89)
R2 = 0.94 a = 0.09 Sc = -2.21 DW = 1.74
4. Other oil revenues:
Ln (
OTHOIL
0.659+0.288 Ln( GDP*PGNP )+
* PGNP11
(2.11) (4.59)
* PGNPt-1
0.1012 Liz ( )+O.0397 Ln DPOIL—t
OTHOILt-1
(1.33) (6.01)
R2 = 0.68 a = 1.63 DW = -2.07 sc = 1.56

5. Corporation or company tax:
COMTAX GNP * PGNP GNP PGNP)

*
Ln( -0.260+0. 168Ln( )+O.O44Ln(
COMTAX
(1.08) (3.34) (1.09)
R2 = 0.68 a= 0.23 SC = 1.45 DW = -2.31
6. Manufactures exports:
Ln (
MAEXP
+0.4464 Ln
0 GDP
(
0 GDP:1) REXR (PXM

+0.1568 Ln ( + 0.189 Ln )
PGNP
(2.02) (1.48) (1.87)
R2 = 0.71 a = 0.24 DW = 2.17 SC = -1.94

7. Primary commodities:
PMEXP GDP
Ln ( ) —-0.1574 + 0.2635 Ln ( ) +
(-0.88) (1.88)
GDP REXR PCOM

0.4837 La ( +0.109 Ln ( +0.148 Ln (
GDP11 PGNP
(2.78) (1.68) (2.19)
R2 = 0.69 a = 0.2 DW = 2.13 SC = -2.61
8. Imports of consumer goods:
(
Ln ) —0.193 + 0.257 La ) -0.4229 La )—
(1.12) (2.52) (2.16)

78 RESEARCN PAPER 32
0.168 Ln (
REXR
) -i-O.2O6Ln (
1'D tl )
CNJMP11
(1.93) (2.09)
R2 = 0.77 = 0.16 DW = 1.78 SC = -3.75
9. Imports of capital goods and raw materials:
CA iMP
Ln --0.2276 +0.545 L n( GDP )+
GDP11
(2.85)
GDP,1
0.382 Ln ( -0.4108 Ln (REXR) -0.241
(3.08) (2.01) (1.89)
R2 = 0.89 a = 0.106 DW = 1.87 SC = -4.14

10. Money demand equation:
Ln
(M).-Ø•1074 + 0.1506 Ln (GNP)+0.7889 Ln (P'NP)tl_0.OO54DRS
(1.18) (2.32) (2.48) (-2.03)
R2 = 0.94 a = 0.09 DW = 1.97 SC = -4.45
11. Monetary policy reaction function (short-term nominal rate):
0.333 Ln (f.?)
+ M
DRS -DRS ÷0.1044 (DRS +1
-DRS 11
—0.0054
12. Nominal long-term rate (adaptive expectations):
LR -1
13. capacity output:
Ln(YCAP) = 0.7946 + 0.3478Ln(KP) +0.6522Ln(LAB) +
(2.02) (3.79) (4.46)
0.2092Ln(KG) + 0.0371T
(2.26) (1.73)
R2 = 0.93; Restricted to constant returns (sum of coefficients on KP and LAB

equals unity).
14. Labour demand:
*LFD- -0.1034+0.2354 A CAPUTL-0.2005 A WG+0.6982 1
(-0.27) (3.13) (-1.87) (2.18)
R2 = 0.69 a = 0.22 DW = 2.09 SC = -3.88
15. Wage rate:
Ln(
WG )—-0.0149+0.0243
(INFe
0624 Lii
GDP
)-0.O33AUER
(
jjtJJ4'
(-2.54) (1.92) (2.68) (-2.02)
R2 = 0.82 a = 0.064 DW = 2.23 Sc = -5.12
16. Inflation rate:
INF WG*PGNP DPIT +INF,1

-—0.0055 +0.1853 Lii ( )+O.385 Lii (
PGNP T
1 1 -1
(2.09) (2.26)
CAPTUL
0.2259 Ln ( ) +0.583
CAPTUL,.1
(2.12) (1.93)
R2 = 0.88 a = 0.11 DW = 2.03 SC -3.69
17. Voluntary or actual external debt service payments (reaction

function):
VOLSEV-0. I2OYCAP+0.48(A YCAP)+0. * PDEBT, *
18. New flow of private external lending to the private sector:
£NPCAPO.740+0. 1 14A(TEXP-OILEXP)+0.0418A
(2.94) (1.05) (1.62) (2.38)
R2 = 0.63 a = 0.29 sc = -1.78

Appendix 4. Simulations results figures

Figure 1-A: Fiscal contraction shock
ci)
C
a)
(I)
cci
E
0
C
0
cci
>
ci)
0
1993 1995 2000 2004

Year of simulation
Figure 1-B: Fiscal contraction shock contd
cD
C
(I,
E
0
C
0
>
a,
a
Primary commodity exports (%)
Year of simulation
Figure 1-C: Fiscal contraction shock... .contd
Manufacturer exports (%) Short-term nominal interest rate

0.16
—.ADAP 0.2'
-- -. CONS
0.12
-0.2
0.08W -0.4
v'.'
-0.6
/'
1
II
7
-0.8 /
—0.02
0.4
•1
I I
Consumer goods imports

I I I I
-1.0
—1.2
0.04'
i'ti/ I I
Long-term real interest rate

I I I I I I
(1) 0.02
0.2
1
-0.2 -0.04
/
." / -0.06
u.d
' I I
0.12
•" l I I I I I I I
Capital goods imports (%)

1.0 0.02 Short-term real interest rate
-1.0
C
/
—2.0 ' . _• —— •, —
-0.06.
-3.0
r
.
-0.08 I I I
1993 1995 2000 2004 1993 1995 2000 2004

Figure 1-D: Fiscal contraction shock . .. .contd
Nominal exchange rate .—.----

.—.——-
__.
_.— _.
s__I—
C
a,
'I,
Ca
.0
E
2
C
0
Ca
>
a,
External debt service payments (%)
1993 1995 2000 2004

Year of simulation
Figure 2-A: Oil price shock (model-consistent expectations)
Real GDP (%) Price level (%)
a) Private investment (%)

C
ci)
(0
a)
E
0
C
0
cci
>
a)
Total exports (%)
1993 1995 2002000 1993 1995 2000 2004

Year of simulation
_______________________________
Figure 2-B: Oil price shock (model-consistent expectations)... .contd
n .r
0.07
9.65
0.066
0.062
0.058
9.45
9351 1111111111 0.054

0.0521 IIlliilliiI
Primary commodity exports (%) Short-term nominal interest rate
a)
0.028 —0.22 -
U)
0.024
2
0.02
5
-0.3
0.016
0.014' I
_U.i11 1111111 Ill
Real long-term interest rate
Default risk premium
-0.0024 . -0.012
-0.0028 —0.016
-0.02
-0.0032
- 0.024
—0.0036
—0.004 '
1993 1995 2000 2004
— -0.03
1993 1995 2000 2004
Year of simulation
88 //III/IflIffhJIJJiiJfli/ffiuiiiijiii RESEARCH PAPER 32
Figure 2-C: Oil price shock (model-consistent expectations)... .contd
Income tax revenue (%)
Total government revenue (%)

External debt service payments (%)
C)
C
a)
U)
CC
.0
E
0
C
0
CC
>
a)
a
Trade balance (Nm)

External loans to private sector (%)
100
90
80
Year of simulation
AFRICAN ECONOMIC RESEARCH CONSORTIUM
The principal objective of the African Economic Research

Consortium (AERC), established in August 1988, is to
strengthen local capacity for conducting independent, rigorous
inquiry into problems pertinent to the management of
economies in Sub-Saharan Africa.
In response to special needs of the region, AERC has
P.O. BOX 62882 adopted a flexible approach to improve the technical skills of
NAIROBI, KENYA local researchers, allow for regional determination of research
priorities, strengthen national institutions concerned with
economic policy research, and facilitate closer ties between
researchers and policy makers.
Since its establishment, AERC has been supported by
TELEPHONE (254-2) 228057 private foundations, bilateral aid agencies and international
225234 215898 212359
organizations.
332438 225087
SPECIAL PAPERS contain the findings of commissioned
studies in furtherance of AERCs programmes for research,
training and capacity building.
RESEARCH PAPERS contain the edited and externally
TELEX 22480 reviewed results of research financed by the AERC.
It is AERC's policy that authors of Special and Research
Papers are free to use material contained therein in other
publications. Views expressed in the Special and Research
Papers are those of the authors alone and should not be
FAX (254-2) 219308
attributed to the AERC's sponsoring Members, Advisory
Committee, or Secretariat.
Further information concerning the AERC and additional
copies of Special and Research Papers can be obtained by
writing to: African Economic Research Consortium, P.O. Box
62882, Nairobi, Kenya.
ISBN 9966-900-26-8

Macroeconomic Adjustment, Trade Growth: Policy Macroeconomic Model Nigeria

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Macroeconomic Adjustment, Trade Growth: Policy Macroeconomic Model Nigeria

Uploaded by

Copyright:

Available Formats

MARCH 1995

RESEARCH PAPER THIRTY-TWO

Macroeconomic adjustment, trade

© 1994, African Economic Research Consortium.

Structural Adjustment Programmes and the Coffee Sector in Uganda by Germina

The Informal Financial Sector and Macroeconomic Adjustment in Malawi by C. Chipeta

The Effects of Non-Bank Financial Intermediaries on Demand for Money in Kenya by

A Macroeconomic-Demographic Modelfor Ethiopia by Asmerom Kidane, Research Paper

Macroeconomic Approach to External Debt: the Case of Nigeria by S. Ibi Ajayi,

Financial System Regulation, Deregulation and Savings Mobilization in Nigeria by A.

Revenue Productivity Implications of Tax Reform in Tanzania by Nehemiah E. Osoro, Research

Macroeconomic Constraints and Medium-Term Growth in Kenya: A Three-Gap Analysis by F.M.

Constraints to the Development of Non-Traditional Exports in Uganda, 1 981-90

Monetary Harmonization in Southern Africa by C. Chipeta and M.L.C. Mukandwire, Research

Charles Chukwuma Soludo

AERC Research Paper 32

Typeset by the African Economic Research Consortium,

Printed by The Regal Press Kenya Ltd.,

Appendix 1: Data Sources and Limitations 69

Introduction/statement of the problem

2. In view of the institutional practice of active monetization of foreign exchange

4. In the absence of debt forgiveness or equivalent operations, and in the face of

Objectives of the research agenda

The specific objectives of the present research are:

1. to build a medium-sized macroeconomic model of the Nigerian economy that

about the expectations formation process of private economic agents in Nigeria

3. to make policy recommendations on the basis of the simulation results.

Overview of major macroeconomic trends during the

• abolition of the import-licensing regime and the liberalization of the foreign

• rationalization and restructurLg of tariffs in order to aid industrial diversification;

• introduction of measures to stimulate domestic production and broaden the supply

• substantial liberalization of trade and payments;

• adoption of appropriate pricing policies, especially for petroleum products and

• encouragement of rationalization and privatization of public sector enterprises.

It is one thing to articulate sound economic policies; it is another to effectively

Table 1: Some economic and financial indicators

CBN's Statistical Bulletin, 1992

GDP = Growth rate of GDP

The larger deficit, in terms, was due mainly to large external

Table 2: Exports by major commodity groups

Agricultural Mineral Manufacture! Textiles Chemicals Other Other

1986 4.6 93.8 0.6 1.0

Table 3: Exports by major trading partners (percentage of total exports)

Common Eastern Japan UK USA - ECOWAS Western Others

1985 5.99 0.28 0.06 4.60 18.06 3.31 61.58 6.12

Source: Central Bank of Nigeria's: Statistical Bulletin, 1992.

Table 4: Imports by major groups (jercentage of total imports)

1985 1986 1987 1988 1989 1990 1991

Clarification of methodological issues

controlled by the government. Also, economic policy in these environments is perceived

Summary of the model structure

Figure 1. Flow chart showing major linkages in the model blocks.

Block A model specifications: domestic absorption

Domestic absorption comprises private consumption expenditures (CP), private

Private consumption expenditures

= a0 +allnYD1 +aIn4RLENR (1.1)

YD, YTAX + DSERV + (URS*PFA,,)1ER -

YD, = CP, -l-INV, + ((MT, +(PFA,-PFA,,) -

disposable income is allocated to consumption, investment expenditures and net

Private investment expenditures

The evolution of the capital stock is given by

where Dep denotes the rate of depreciation.

KP, = +INV1 - Dep*(KPt1) (2.3a)

(3) CUSTAX = tarifMIMPPGNP + OILEXP)*PGNP (3.4)